=== This is the CP2K Performance-Test === Updating 91e7af0a0..8c0411e32 Fast-forward arch/Linux-gnu-aarch64-static.ssmp | 3 ++- arch/Linux-gnu-aarch64.psmp | 3 ++- arch/Linux-gnu-x86_64-static.ssmp | 3 ++- arch/Linux-gnu-x86_64.psmp | 3 ++- 4 files changed, 8 insertions(+), 4 deletions(-) Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: 8c0411e32bbb6770871f8515127d8b2d742a8838 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2021.11.002, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 5.2.3, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.0, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (15.09.2022) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2021.11.002 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 5.2.3 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := b4336484fb65b0e73211a8f920ae4361c7c353fd USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/01 job id: 41846573 --- Point --- name: 10 plot: h2o_32_ri_rpa_mp2 regex: Total RI-RPA Time= label: RI-RPA (8n/2r/6t) --- Point --- name: 11 plot: h2o_32_ri_rpa_mp2_mem regex: Estimated peak process memory label: RI-RPA (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-MP2.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-HF-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-MP2-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/02 job id: 41846574 --- Point --- name: 20 plot: h2o_32_ri_rpa_mp2 regex: Total MP2 Time= label: RI-MP2 (8n/6r/2t) --- Point --- name: 21 plot: h2o_32_ri_rpa_mp2_mem regex: Estimated peak process memory label: RI-MP2 (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/03 job id: 41846576 --- Point --- name: 100 plot: h2o_64_md regex: CP2K label: (8n/12r/1t) --- Point --- name: 101 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/04 job id: 41846577 --- Point --- name: 102 plot: h2o_64_md regex: CP2K label: (8n/6r/2t) --- Point --- name: 103 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/05 job id: 41846579 --- Point --- name: 104 plot: h2o_64_md regex: CP2K label: (8n/4r/3t) --- Point --- name: 105 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/06 job id: 41846580 --- Point --- name: 106 plot: h2o_64_md regex: CP2K label: (8n/3r/4t) --- Point --- name: 107 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/07 job id: 41846582 --- Point --- name: 108 plot: h2o_64_md regex: CP2K label: (8n/2r/6t) --- Point --- name: 109 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/08 job id: 41846584 --- Point --- name: 110 plot: h2o_64_md regex: CP2K label: (8n/1r/12t) --- Point --- name: 111 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-128 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-128.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/09 job id: 41846586 --- Point --- name: 200 plot: h2o_128_md regex: CP2K label: (8n/12r/1t) --- Point --- name: 201 plot: h2o_128_md_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-128 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-128.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/10 job id: 41846587 --- Point --- name: 202 plot: h2o_128_md regex: CP2K label: (8n/6r/2t) --- Point --- name: 203 plot: h2o_128_md_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-128 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-128.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/11 job id: 41846589 --- Point --- name: 204 plot: h2o_128_md regex: CP2K label: (8n/4r/3t) --- Point --- name: 205 plot: h2o_128_md_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-128 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-128.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/12 job id: 41846591 --- Point --- name: 206 plot: h2o_128_md regex: CP2K label: (8n/3r/4t) --- Point --- name: 207 plot: h2o_128_md_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-128 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-128.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/13 job id: 41846593 --- Point --- name: 208 plot: h2o_128_md regex: CP2K label: (8n/2r/6t) --- Point --- name: 209 plot: h2o_128_md_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-128 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-128.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/14 job id: 41846595 --- Point --- name: 210 plot: h2o_128_md regex: CP2K label: (8n/1r/12t) --- Point --- name: 211 plot: h2o_128_md_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-256 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-256.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 30 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/15 job id: 41846599 --- Point --- name: 400 plot: h2o_256_md regex: CP2K label: (8n/12r/1t) --- Point --- name: 401 plot: h2o_256_md_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-256 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-256.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 30 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/16 job id: 41846600 --- Point --- name: 402 plot: h2o_256_md regex: CP2K label: (8n/6r/2t) --- Point --- name: 403 plot: h2o_256_md_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-256 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-256.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 30 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/17 job id: 41846604 --- Point --- name: 404 plot: h2o_256_md regex: CP2K label: (8n/4r/3t) --- Point --- name: 405 plot: h2o_256_md_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-256 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-256.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 30 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/18 job id: 41846607 --- Point --- name: 406 plot: h2o_256_md regex: CP2K label: (8n/3r/4t) --- Point --- name: 407 plot: h2o_256_md_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-256 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-256.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 30 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/19 job id: 41846609 --- Point --- name: 408 plot: h2o_256_md regex: CP2K label: (8n/2r/6t) --- Point --- name: 409 plot: h2o_256_md_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-256 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-256.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 30 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/20 job id: 41846612 --- Point --- name: 410 plot: h2o_256_md regex: CP2K label: (8n/1r/12t) --- Point --- name: 411 plot: h2o_256_md_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/21 job id: 41846618 --- Point --- name: 500 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/12r/1t) --- Point --- name: 501 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/22 job id: 41846621 --- Point --- name: 502 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/6r/2t) --- Point --- name: 503 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/23 job id: 41846625 --- Point --- name: 504 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/4r/3t) --- Point --- name: 505 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/24 job id: 41846628 --- Point --- name: 506 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/3r/4t) --- Point --- name: 507 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/25 job id: 41846632 --- Point --- name: 508 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/2r/6t) --- Point --- name: 509 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/26 job id: 41846633 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.033 133.918 133.919 farming_run 1 2.0 133.246 133.250 133.891 133.893 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 4194304 0.0% 0.0% 100.0% flops 14 x 32 x 32 154140672 0.0% 0.0% 100.0% flops 29 x 32 x 32 159645696 0.0% 0.0% 100.0% flops 14 x 14 x 32 208732160 0.0% 0.0% 100.0% flops 29 x 14 x 32 212860928 0.0% 0.0% 100.0% flops 14 x 29 x 32 212860928 0.0% 0.0% 100.0% flops 29 x 29 x 32 227352576 0.0% 0.0% 100.0% flops 14 x 32 x 14 896801644032 0.0% 0.0% 100.0% flops 29 x 32 x 14 928925089792 0.0% 0.0% 100.0% flops 14 x 32 x 29 928925089792 0.0% 0.0% 100.0% flops 29 x 32 x 29 962100985856 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693169221632 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753639550976 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.164741E+12 0.0% 0.0% 100.0% flops max/rank 447.801317E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249492158 0.0% 0.0% 100.0% number of processed stacks 164328 0.0% 0.0% 100.0% average stack size 0.0 0.0 1518.3 marketing flops 7.165779E+12 ------------------------------------------------------------------------------- # multiplications 1160 max memory usage/rank 1.463988E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 489 2263609. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.027 115.637 115.637 qs_energies 1 2.0 0.000 0.000 115.373 115.376 mp2_main 1 3.0 0.000 0.000 113.535 113.538 mp2_gpw_main 1 4.0 0.372 0.380 112.712 112.715 mp2_ri_gpw_compute_in 1 5.0 0.174 0.179 93.549 93.869 mp2_ri_gpw_compute_in_loop 1 6.0 0.005 0.005 55.753 56.072 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.175 42.094 47.382 get_2c_integrals 1 6.0 0.000 0.000 36.889 37.622 integrate_v_rspace 273 8.0 0.438 0.450 25.311 30.360 pw_transfer 6555 10.6 0.380 0.394 27.435 27.912 grid_integrate_task_list 273 9.0 21.097 26.661 21.097 26.661 fft_wrap_pw1pw2 5465 11.4 0.046 0.048 26.113 26.586 fft_wrap_pw1pw2_100 2178 12.4 1.174 1.333 23.679 24.157 compute_2c_integrals 1 7.0 0.011 0.012 19.347 19.347 rpa_ri_compute_en 1 5.0 0.000 0.000 18.699 19.062 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.881 19.022 mp2_eri_2c_integrate_gpw 1 9.0 2.400 2.439 18.878 19.020 cp_fm_cholesky_decompose 12 8.2 17.500 18.246 17.500 18.246 cholesky_decomp 1 7.0 0.000 0.000 16.388 17.125 fft3d_s 5443 13.4 16.234 16.658 16.255 16.677 ao_to_mo_and_store_B_mult_1 272 7.0 10.854 15.588 10.854 15.588 calculate_wavefunction 272 8.0 5.517 5.660 12.695 13.264 rpa_num_int 1 6.0 0.000 0.000 10.598 10.608 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.515 10.522 calc_mat_Q 8 8.0 0.000 0.000 9.354 9.456 contract_S_to_Q 8 9.0 0.000 0.000 8.774 8.878 calc_potential_gpw 544 9.5 0.005 0.006 8.240 8.600 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.002 0.002 8.205 8.439 parallel_gemm_fm 14 9.1 0.000 0.000 8.367 8.437 parallel_gemm_fm_cosma 14 10.1 8.367 8.437 8.367 8.437 potential_pw2rs 545 10.0 0.108 0.111 7.724 8.385 create_integ_mat 1 6.0 0.014 0.028 7.687 7.697 collocate_single_gaussian 272 10.0 0.040 0.041 7.451 7.650 array2fm 1 7.0 0.000 0.000 6.656 7.168 pw_scatter_s 2720 13.7 4.461 4.683 4.461 4.683 pw_gather_s 2722 13.2 3.882 4.197 3.882 4.197 array2fm_buffer_send 1 8.0 2.954 3.145 2.954 3.145 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.715307, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2735.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.063 402.492 402.494 farming_run 1 2.0 401.623 401.643 402.429 402.434 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.223098E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1825 23678. MP_Sync 38 MP_Alltoall 77 13578446. MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.051 210.626 210.627 qs_energies 1 2.0 0.000 0.000 210.326 210.337 scf_env_do_scf 1 3.0 0.000 0.000 106.350 106.350 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.514 105.521 rebuild_ks_matrix 4 6.0 0.000 0.000 105.512 105.519 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.061 105.512 105.519 hfx_ks_matrix 4 8.0 0.001 0.001 105.149 105.152 integrate_four_center 4 9.0 0.144 0.469 105.148 105.151 mp2_main 1 3.0 0.000 0.000 103.692 103.703 mp2_gpw_main 1 4.0 0.037 0.228 102.839 102.852 integrate_four_center_main 4 10.0 0.071 0.595 96.832 99.484 integrate_four_center_bin 267 11.0 96.761 99.481 96.761 99.481 init_scf_loop 1 4.0 0.000 0.000 92.239 92.241 mp2_ri_gpw_compute_in 1 5.0 0.070 0.318 75.471 77.071 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 55.077 56.681 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.160 42.050 47.309 integrate_v_rspace 95 8.0 0.397 0.562 28.377 33.415 pw_transfer 2240 10.6 0.144 0.169 29.928 30.272 mp2_ri_gpw_compute_en 1 5.0 0.060 0.072 27.215 29.584 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.935 29.300 grid_integrate_task_list 95 9.0 23.667 28.925 23.667 28.925 fft_wrap_pw1pw2_100 730 12.4 1.280 1.414 26.680 27.029 ao_to_mo_and_store_B_mult_1 91 7.0 11.337 26.799 11.337 26.799 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.827 1.895 24.946 24.957 get_2c_integrals 1 6.0 0.000 0.002 20.301 20.324 compute_2c_integrals 1 7.0 0.003 0.004 19.277 19.281 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.913 19.152 mp2_eri_2c_integrate_gpw 1 9.0 1.740 1.845 18.912 19.150 fft3d_s 1823 13.4 18.424 18.747 18.437 18.760 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.109 14.110 calculate_wavefunction 91 8.0 2.052 2.212 9.775 10.002 mp2_ri_gpw_compute_en_expansio 172 7.0 0.559 0.597 8.727 9.428 potential_pw2rs 186 10.0 0.034 0.035 8.657 9.131 local_gemm 172 8.0 8.168 8.851 8.168 8.851 mp2_ri_gpw_compute_en_comm 22 7.0 0.513 0.537 8.004 8.650 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.266 8.588 collocate_single_gaussian 91 10.0 0.017 0.025 7.907 8.161 calc_potential_gpw 182 9.5 0.002 0.002 7.919 8.131 mp_sendrecv_dm3 2068 8.0 5.999 6.641 5.999 6.641 mp2_ri_gpw_compute_en_ener 172 7.0 6.344 6.441 6.344 6.441 mp_sync 38 10.4 2.989 5.753 2.989 5.753 pw_gather_s 912 13.2 4.942 5.363 4.942 5.363 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.829703, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1510.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 449.937408E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 MP_ISendRecv 45220 5520. MP_Wait 60486 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.057 53.490 53.491 qs_mol_dyn_low 1 2.0 0.003 0.003 53.153 53.161 qs_forces 11 3.9 0.002 0.002 53.087 53.087 qs_energies 11 4.9 0.022 0.171 51.615 51.626 scf_env_do_scf 11 5.9 0.001 0.005 45.139 45.161 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 42.777 42.778 dbcsr_multiply_generic 2286 12.5 0.091 0.096 34.071 34.515 qs_scf_new_mos 108 7.5 0.000 0.001 32.638 32.920 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.637 32.920 ot_scf_mini 108 9.5 0.002 0.002 30.960 31.155 multiply_cannon 2286 13.5 0.185 0.191 26.564 27.984 multiply_cannon_loop 2286 14.5 1.494 1.561 25.818 27.266 velocity_verlet 10 3.0 0.001 0.002 25.996 25.996 ot_mini 108 10.5 0.001 0.001 19.731 19.960 qs_ot_get_derivative 108 11.5 0.001 0.001 16.764 16.962 mp_waitall_1 267858 16.1 9.119 15.196 9.119 15.196 multiply_cannon_metrocomm3 54864 15.5 0.066 0.070 6.222 13.383 multiply_cannon_multrec 54864 15.5 4.173 6.487 7.896 11.389 rebuild_ks_matrix 119 8.3 0.000 0.000 7.990 8.119 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.990 8.118 multiply_cannon_sync_h2d 54864 15.5 5.814 7.638 5.814 7.638 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.058 7.173 mp_sum_l 7207 12.9 5.433 7.050 5.433 7.050 qs_ot_get_p 119 10.4 0.001 0.001 6.510 6.820 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.924 6.391 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.463 5.601 dbcsr_mm_accdrv_process 76910 16.1 1.180 1.838 3.646 5.141 init_scf_run 11 5.9 0.000 0.001 4.819 4.819 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.819 4.819 sum_up_and_integrate 119 10.3 0.013 0.016 4.550 4.557 integrate_v_rspace 119 11.3 0.002 0.003 4.537 4.546 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.184 4.269 calculate_rho_elec 119 8.7 0.011 0.016 4.183 4.269 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.522 3.570 multiply_cannon_metrocomm1 54864 15.5 0.052 0.058 1.899 3.304 jit_kernel_multiply 13 15.8 2.403 3.216 2.403 3.216 calculate_dm_sparse 119 9.5 0.000 0.000 3.058 3.178 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.889 3.095 apply_single 119 13.6 0.000 0.000 2.889 3.094 rs_pw_transfer 974 11.9 0.012 0.013 2.869 2.939 ot_diis_step 108 11.5 0.006 0.006 2.691 2.691 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.683 2.684 calculate_first_density_matrix 1 7.0 0.000 0.000 2.646 2.650 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.532 2.590 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.431 2.432 cp_fm_redistribute_end 50 14.0 2.203 2.403 2.208 2.405 density_rs2pw 119 9.7 0.004 0.004 2.269 2.391 init_scf_loop 11 6.9 0.000 0.000 2.343 2.366 cp_fm_diag_elpa_base 50 14.0 0.195 2.352 0.196 2.365 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.290 2.293 acc_transpose_blocks 54864 15.5 0.228 0.248 1.793 2.224 grid_integrate_task_list 119 12.3 2.025 2.119 2.025 2.119 wfi_extrapolate 11 7.9 0.001 0.001 2.112 2.112 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.035 2.082 mp_sum_d 4129 12.0 1.429 2.081 1.429 2.081 potential_pw2rs 119 12.3 0.004 0.005 1.933 1.948 make_m2s 4572 13.5 0.053 0.056 1.769 1.816 pw_transfer 1439 11.6 0.052 0.055 1.717 1.791 make_images 4572 14.5 0.132 0.138 1.687 1.733 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.642 1.717 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.549 1.576 fft3d_ps 1201 14.6 0.354 0.456 1.421 1.487 mp_alltoall_d11v 2130 13.8 1.339 1.454 1.339 1.454 grid_collocate_task_list 119 9.7 1.291 1.352 1.291 1.352 parallel_gemm_fm 81 9.0 0.000 0.000 1.339 1.344 parallel_gemm_fm_cosma 81 10.0 1.339 1.343 1.339 1.343 fft_wrap_pw1pw2_140 487 13.2 0.079 0.093 1.250 1.325 mp_waitany 12084 13.8 1.201 1.281 1.201 1.281 mp_alltoall_i22 627 13.8 0.868 1.104 0.868 1.104 dbcsr_dot_sd 1205 11.9 0.047 0.058 0.734 1.098 acc_transpose_blocks_kernels 54864 16.5 0.252 0.387 0.804 1.073 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.491000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=429.000000, yerr=1.414214 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 485.683200E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10226 305. MP_Sync 104 MP_Alltoall 2060 1308042. MP_ISendRecv 33558 37093. MP_Wait 40318 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.029 38.421 38.422 qs_mol_dyn_low 1 2.0 0.003 0.003 38.232 38.241 qs_forces 11 3.9 0.002 0.002 38.170 38.171 qs_energies 11 4.9 0.001 0.001 36.490 36.493 scf_env_do_scf 11 5.9 0.000 0.001 31.035 31.036 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 28.545 28.546 dbcsr_multiply_generic 2286 12.5 0.099 0.103 21.269 21.633 qs_scf_new_mos 108 7.5 0.001 0.001 19.735 19.974 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.734 19.974 ot_scf_mini 108 9.5 0.002 0.003 18.830 19.000 velocity_verlet 10 3.0 0.001 0.002 18.047 18.048 multiply_cannon 2286 13.5 0.209 0.219 16.264 17.908 multiply_cannon_loop 2286 14.5 0.900 0.975 15.145 16.555 ot_mini 108 10.5 0.001 0.001 11.727 11.961 mp_waitall_1 217478 16.2 6.000 11.134 6.000 11.134 multiply_cannon_metrocomm3 27432 15.5 0.068 0.071 4.154 9.521 qs_ot_get_derivative 108 11.5 0.001 0.001 9.263 9.435 multiply_cannon_multrec 27432 15.5 1.971 4.314 5.838 8.775 rebuild_ks_matrix 119 8.3 0.000 0.000 7.036 7.170 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.035 7.169 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.203 6.324 dbcsr_mm_accdrv_process 47894 16.0 2.985 5.235 3.799 5.617 qs_ot_get_p 119 10.4 0.001 0.001 4.191 4.422 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.570 4.415 init_scf_run 11 5.9 0.000 0.001 4.246 4.247 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.246 4.246 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.052 4.131 apply_single 119 13.6 0.000 0.000 3.052 4.131 sum_up_and_integrate 119 10.3 0.025 0.028 4.097 4.104 integrate_v_rspace 119 11.3 0.002 0.003 4.072 4.080 mp_sum_l 7207 12.9 2.042 3.998 2.042 3.998 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.521 3.555 calculate_rho_elec 119 8.7 0.021 0.024 3.521 3.554 multiply_cannon_sync_h2d 27432 15.5 2.189 2.916 2.189 2.916 qs_ot_p2m_diag 50 11.0 0.008 0.013 2.755 2.774 calculate_first_density_matrix 1 7.0 0.000 0.000 2.770 2.773 make_m2s 4572 13.5 0.052 0.054 2.511 2.744 make_images 4572 14.5 0.200 0.239 2.425 2.654 rs_pw_transfer 974 11.9 0.010 0.011 2.500 2.591 init_scf_loop 11 6.9 0.000 0.000 2.470 2.470 ot_diis_step 108 11.5 0.011 0.011 2.411 2.412 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.323 2.324 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.134 2.222 calculate_dm_sparse 119 9.5 0.000 0.000 2.050 2.126 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.988 1.989 density_rs2pw 119 9.7 0.004 0.004 1.895 1.979 cp_fm_redistribute_end 50 14.0 1.639 1.963 1.643 1.964 cp_fm_diag_elpa_base 50 14.0 0.311 1.903 0.319 1.934 grid_integrate_task_list 119 12.3 1.831 1.917 1.831 1.917 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.889 1.892 potential_pw2rs 119 12.3 0.006 0.007 1.854 1.862 jit_kernel_multiply 9 16.1 0.761 1.733 0.761 1.733 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.689 1.731 pw_transfer 1439 11.6 0.064 0.067 1.667 1.699 prepare_preconditioner 11 7.9 0.000 0.000 1.586 1.613 make_preconditioner 11 8.9 0.000 0.000 1.586 1.613 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.577 1.610 make_images_data 4572 15.5 0.045 0.051 1.158 1.576 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.491 1.548 acc_transpose_blocks 27432 15.5 0.107 0.110 1.192 1.506 hybrid_alltoall_any 4725 16.4 0.050 0.111 1.005 1.476 wfi_extrapolate 11 7.9 0.001 0.001 1.428 1.428 mp_allgather_i34 2286 14.5 0.560 1.331 0.560 1.331 mp_alltoall_d11v 2130 13.8 1.175 1.316 1.175 1.316 fft3d_ps 1201 14.6 0.492 0.544 1.283 1.312 grid_collocate_task_list 119 9.7 1.242 1.306 1.242 1.306 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.257 1.264 fft_wrap_pw1pw2_140 487 13.2 0.076 0.083 1.213 1.244 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.182 1.232 mp_sum_d 4129 12.0 0.556 0.998 0.556 0.998 qs_energies_init_hamiltonians 11 5.9 0.022 0.029 0.945 0.945 parallel_gemm_fm 81 9.0 0.000 0.000 0.932 0.937 parallel_gemm_fm_cosma 81 10.0 0.932 0.937 0.932 0.937 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.888 0.901 acc_transpose_blocks_kernels 27432 16.5 0.182 0.271 0.665 0.886 make_basis_sm 11 9.8 0.000 0.000 0.836 0.838 rs_pw_transfer_PW2RS_50 119 14.3 0.589 0.607 0.747 0.778 make_images_sizes 4572 15.5 0.005 0.005 0.535 0.770 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.422000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=462.181818, yerr=1.192262 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 520.065024E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_ISendRecv 22134 57667. MP_Wait 33054 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.040 35.824 35.825 qs_mol_dyn_low 1 2.0 0.004 0.006 35.356 35.364 qs_forces 11 3.9 0.002 0.004 35.284 35.285 qs_energies 11 4.9 0.001 0.003 33.686 33.688 scf_env_do_scf 11 5.9 0.001 0.001 26.106 26.106 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.014 23.014 dbcsr_multiply_generic 2286 12.5 0.095 0.110 19.220 19.304 multiply_cannon 2286 13.5 0.196 0.204 15.748 16.476 multiply_cannon_loop 2286 14.5 0.639 0.667 14.934 15.681 velocity_verlet 10 3.0 0.002 0.008 15.442 15.443 qs_scf_new_mos 108 7.5 0.001 0.001 14.874 14.904 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.873 14.904 ot_scf_mini 108 9.5 0.002 0.004 14.144 14.157 multiply_cannon_multrec 18288 15.5 1.953 2.893 9.465 9.730 ot_mini 108 10.5 0.001 0.001 8.810 8.828 dbcsr_mm_accdrv_process 38222 16.0 5.865 8.168 7.429 8.311 qs_ot_get_derivative 108 11.5 0.001 0.001 7.303 7.319 rebuild_ks_matrix 119 8.3 0.000 0.000 6.431 6.446 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 6.431 6.446 init_scf_run 11 5.9 0.001 0.004 6.334 6.334 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.333 6.334 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.682 5.694 calculate_first_density_matrix 1 7.0 0.000 0.001 5.063 5.064 jit_kernel_multiply 10 16.0 1.512 4.594 1.512 4.594 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 4.471 4.473 mp_waitall_1 169478 16.3 3.087 4.324 3.087 4.324 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 4.253 4.259 sum_up_and_integrate 119 10.3 0.031 0.032 4.043 4.047 integrate_v_rspace 119 11.3 0.002 0.003 4.012 4.020 qs_ot_get_p 119 10.4 0.001 0.001 3.619 3.640 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.718 3.349 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.320 3.324 calculate_rho_elec 119 8.7 0.030 0.031 3.319 3.323 init_scf_loop 11 6.9 0.000 0.002 3.074 3.076 rs_pw_transfer 974 11.9 0.009 0.010 2.446 2.514 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.369 2.414 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.051 2.367 apply_single 119 13.6 0.000 0.000 2.051 2.367 make_m2s 4572 13.5 0.044 0.045 1.969 2.134 make_images 4572 14.5 0.191 0.205 1.885 2.046 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.039 2.044 density_rs2pw 119 9.7 0.004 0.004 1.906 2.029 calculate_dm_sparse 119 9.5 0.000 0.000 1.882 1.904 grid_integrate_task_list 119 12.3 1.794 1.880 1.794 1.880 potential_pw2rs 119 12.3 0.007 0.008 1.811 1.824 prepare_preconditioner 11 7.9 0.000 0.000 1.809 1.811 make_preconditioner 11 8.9 0.000 0.000 1.809 1.811 pw_transfer 1439 11.6 0.063 0.067 1.776 1.786 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.774 1.775 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.655 1.739 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.685 1.697 multiply_cannon_sync_h2d 18288 15.5 1.402 1.631 1.402 1.631 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.603 1.610 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.536 1.537 cp_fm_redistribute_end 50 14.0 1.142 1.515 1.143 1.516 cp_fm_diag_elpa_base 50 14.0 0.357 1.449 0.371 1.487 ot_diis_step 108 11.5 0.011 0.014 1.487 1.487 mp_sum_l 7207 12.9 1.116 1.448 1.116 1.448 fft3d_ps 1201 14.6 0.509 0.526 1.372 1.384 fft_wrap_pw1pw2_140 487 13.2 0.086 0.089 1.318 1.329 grid_collocate_task_list 119 9.7 1.204 1.317 1.204 1.317 acc_transpose_blocks 18288 15.5 0.074 0.076 1.239 1.253 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.244 1.250 wfi_extrapolate 11 7.9 0.001 0.001 1.222 1.223 make_images_data 4572 15.5 0.045 0.049 0.866 1.060 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.016 1.017 arnoldi_extremal 119 11.4 0.050 0.391 0.917 0.952 arnoldi_normal_ev 119 12.4 0.213 0.404 0.867 0.951 hybrid_alltoall_any 4725 16.4 0.055 0.115 0.727 0.944 mp_alltoall_d11v 2130 13.8 0.743 0.911 0.743 0.911 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.887 0.910 mp_sum_d 4129 12.0 0.560 0.867 0.560 0.867 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.820 0.822 acc_transpose_blocks_kernels 18288 16.5 0.211 0.219 0.801 0.812 cp_fm_cholesky_invert 11 10.9 0.775 0.779 0.775 0.779 mp_alltoall_z22v 1201 16.6 0.704 0.770 0.704 0.770 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.648 0.757 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=35.825000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=494.454545, yerr=1.827250 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 554.262528E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_ISendRecv 16422 74133. MP_Wait 24482 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.053 0.147 37.643 37.644 qs_mol_dyn_low 1 2.0 0.005 0.010 37.058 37.065 qs_forces 11 3.9 0.003 0.009 36.986 36.987 qs_energies 11 4.9 0.002 0.005 35.226 35.232 scf_env_do_scf 11 5.9 0.001 0.002 29.353 29.354 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.801 25.803 dbcsr_multiply_generic 2286 12.5 0.100 0.103 19.281 19.473 velocity_verlet 10 3.0 0.002 0.002 18.611 18.613 qs_scf_new_mos 108 7.5 0.001 0.001 17.182 17.233 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.181 17.232 ot_scf_mini 108 9.5 0.003 0.004 16.236 16.286 multiply_cannon 2286 13.5 0.230 0.263 15.060 15.473 multiply_cannon_loop 2286 14.5 0.935 0.965 14.044 14.507 ot_mini 108 10.5 0.005 0.039 10.141 10.202 multiply_cannon_multrec 27432 15.5 2.358 3.041 8.815 9.162 qs_ot_get_derivative 108 11.5 0.001 0.001 8.226 8.277 dbcsr_mm_accdrv_process 47916 15.9 5.209 7.233 6.364 7.552 rebuild_ks_matrix 119 8.3 0.000 0.000 6.854 6.898 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.018 6.853 6.897 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.101 6.139 init_scf_run 11 5.9 0.000 0.001 4.429 4.429 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.428 4.429 sum_up_and_integrate 119 10.3 0.035 0.039 4.053 4.063 integrate_v_rspace 119 11.3 0.003 0.004 4.018 4.029 init_scf_loop 11 6.9 0.001 0.002 3.530 3.532 qs_ot_get_p 119 10.4 0.001 0.001 3.464 3.532 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.435 3.475 calculate_rho_elec 119 8.7 0.040 0.046 3.434 3.475 mp_waitall_1 145218 16.4 2.756 3.421 2.756 3.421 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.793 3.280 calculate_first_density_matrix 1 7.0 0.000 0.001 2.919 2.921 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.199 2.650 apply_single 119 13.6 0.000 0.000 2.199 2.650 prepare_preconditioner 11 7.9 0.000 0.000 2.629 2.638 make_preconditioner 11 8.9 0.000 0.001 2.629 2.638 make_m2s 4572 13.5 0.055 0.065 2.499 2.580 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.231 2.561 make_images 4572 14.5 0.274 0.335 2.390 2.470 rs_pw_transfer 974 11.9 0.009 0.010 2.283 2.406 calculate_dm_sparse 119 9.5 0.000 0.000 2.209 2.253 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.048 2.074 jit_kernel_multiply 10 15.7 1.094 2.032 1.094 2.032 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.009 2.020 density_rs2pw 119 9.7 0.003 0.004 1.898 1.996 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.967 1.968 pw_transfer 1439 11.6 0.063 0.067 1.885 1.915 grid_integrate_task_list 119 12.3 1.825 1.892 1.825 1.892 ot_diis_step 108 11.5 0.015 0.040 1.866 1.872 mp_sum_l 7207 12.9 1.314 1.858 1.314 1.858 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.794 1.828 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.760 1.772 potential_pw2rs 119 12.3 0.009 0.010 1.716 1.722 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.680 1.680 multiply_cannon_metrocomm3 27432 15.5 0.038 0.038 1.016 1.600 acc_transpose_blocks 27432 15.5 0.111 0.113 1.519 1.579 fft3d_ps 1201 14.6 0.535 0.589 1.475 1.499 fft_wrap_pw1pw2_140 487 13.2 0.084 0.091 1.440 1.477 wfi_extrapolate 11 7.9 0.001 0.002 1.461 1.462 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.409 1.410 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.383 1.396 cp_fm_redistribute_end 50 14.0 0.933 1.384 0.935 1.386 cp_fm_diag_elpa_base 50 14.0 0.426 1.301 0.447 1.342 grid_collocate_task_list 119 9.7 1.221 1.325 1.221 1.325 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.188 1.189 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.132 1.150 dbcsr_complete_redistribute 329 12.2 0.120 0.143 0.862 1.138 make_images_data 4572 15.5 0.044 0.049 1.009 1.134 cp_fm_upper_to_full 72 13.5 0.817 1.131 0.817 1.131 multiply_cannon_sync_h2d 27432 15.5 0.990 1.062 0.990 1.062 hybrid_alltoall_any 4725 16.4 0.061 0.151 0.863 1.048 mp_alltoall_d11v 2130 13.8 0.931 1.034 0.931 1.034 acc_transpose_blocks_kernels 27432 16.5 0.270 0.279 0.896 0.956 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.620 0.890 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.796 0.871 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.855 0.861 mp_alltoall_z22v 1201 16.6 0.822 0.847 0.822 0.847 make_basis_sm 11 9.8 0.000 0.000 0.804 0.810 parallel_gemm_fm 81 9.0 0.000 0.000 0.795 0.799 parallel_gemm_fm_cosma 81 10.0 0.795 0.799 0.795 0.799 cp_fm_cholesky_invert 11 10.9 0.781 0.785 0.781 0.785 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.644000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=527.363636, yerr=3.497343 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/07/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 117.977176E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1384136 0.0% 0.0% 100.0% average stack size 0.0 0.0 106.2 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 610.230272E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_ISendRecv 10710 94533. MP_Wait 16690 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.397 0.425 31.031 31.032 qs_mol_dyn_low 1 2.0 0.003 0.003 30.353 30.360 qs_forces 11 3.9 0.002 0.002 30.079 30.079 qs_energies 11 4.9 0.002 0.002 28.351 28.354 scf_env_do_scf 11 5.9 0.001 0.002 23.059 23.059 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 20.266 20.266 velocity_verlet 10 3.0 0.002 0.002 15.239 15.256 dbcsr_multiply_generic 2286 12.5 0.092 0.094 13.492 13.578 qs_scf_new_mos 108 7.5 0.001 0.001 12.158 12.179 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.157 12.178 ot_scf_mini 108 9.5 0.002 0.002 11.459 11.484 multiply_cannon 2286 13.5 0.233 0.246 10.490 10.882 multiply_cannon_loop 2286 14.5 0.332 0.345 9.501 9.711 ot_mini 108 10.5 0.001 0.001 6.404 6.432 multiply_cannon_multrec 9144 15.5 1.587 1.826 6.125 6.363 rebuild_ks_matrix 119 8.3 0.000 0.000 6.234 6.262 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.234 6.262 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.591 5.615 qs_ot_get_derivative 108 11.5 0.001 0.001 5.052 5.074 dbcsr_mm_accdrv_process 12550 15.8 3.432 4.465 4.437 4.535 sum_up_and_integrate 119 10.3 0.038 0.042 3.897 3.904 integrate_v_rspace 119 11.3 0.003 0.003 3.859 3.866 init_scf_run 11 5.9 0.000 0.001 3.843 3.844 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.843 3.844 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.464 3.473 calculate_rho_elec 119 8.7 0.059 0.061 3.463 3.472 qs_ot_get_p 119 10.4 0.001 0.001 3.111 3.139 mp_waitall_1 121218 16.5 2.321 2.875 2.321 2.875 init_scf_loop 11 6.9 0.000 0.000 2.770 2.771 calculate_first_density_matrix 1 7.0 0.000 0.000 2.615 2.616 rs_pw_transfer 974 11.9 0.008 0.009 2.103 2.203 jit_kernel_multiply 10 15.9 0.968 2.151 0.968 2.151 make_m2s 4572 13.5 0.034 0.035 2.012 2.147 make_images 4572 14.5 0.268 0.302 1.924 2.057 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.977 1.979 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.952 1.954 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.944 1.953 density_rs2pw 119 9.7 0.003 0.004 1.854 1.946 grid_integrate_task_list 119 12.3 1.847 1.937 1.847 1.937 prepare_preconditioner 11 7.9 0.000 0.000 1.925 1.929 make_preconditioner 11 8.9 0.000 0.000 1.925 1.929 calculate_dm_sparse 119 9.5 0.000 0.000 1.851 1.869 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.807 1.839 pw_transfer 1439 11.6 0.064 0.066 1.814 1.823 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.774 1.775 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.723 1.731 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.702 1.712 potential_pw2rs 119 12.3 0.010 0.011 1.638 1.644 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.444 1.445 cp_fm_redistribute_end 50 14.0 0.719 1.420 0.720 1.421 cp_fm_diag_elpa_base 50 14.0 0.653 1.341 0.698 1.402 fft3d_ps 1201 14.6 0.539 0.550 1.390 1.400 grid_collocate_task_list 119 9.7 1.275 1.374 1.275 1.374 fft_wrap_pw1pw2_140 487 13.2 0.082 0.086 1.351 1.362 ot_diis_step 108 11.5 0.013 0.013 1.336 1.337 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.287 1.301 acc_transpose_blocks 9144 15.5 0.039 0.040 1.262 1.268 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.257 1.262 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.218 1.219 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.171 1.198 apply_single 119 13.6 0.000 0.000 1.171 1.198 wfi_extrapolate 11 7.9 0.001 0.001 1.182 1.183 hybrid_alltoall_any 4725 16.4 0.062 0.176 0.871 1.143 make_images_data 4572 15.5 0.039 0.042 0.925 1.140 mp_alltoall_d11v 2130 13.8 0.933 1.033 0.933 1.033 acc_transpose_blocks_kernels 9144 16.5 0.117 0.121 1.022 1.026 cp_fm_cholesky_invert 11 10.9 0.982 0.985 0.982 0.985 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.871 0.924 jit_kernel_transpose 5 15.6 0.904 0.907 0.904 0.907 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.773 0.780 multiply_cannon_sync_h2d 9144 15.5 0.708 0.780 0.708 0.780 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.766 0.773 multiply_cannon_metrocomm3 9144 15.5 0.018 0.019 0.387 0.773 mp_allgather_i34 2286 14.5 0.317 0.772 0.317 0.772 mp_alltoall_z22v 1201 16.6 0.724 0.755 0.724 0.755 mp_sum_l 7207 12.9 0.573 0.739 0.573 0.739 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.671 0.720 dbcsr_complete_redistribute 329 12.2 0.193 0.224 0.667 0.703 mp_waitany 5200 13.7 0.518 0.622 0.518 0.622 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=31.032000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=578.090909, yerr=5.071391 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/08/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 235.585836E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1388964 0.0% 0.0% 100.0% average stack size 0.0 0.0 105.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 751.243264E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_ISendRecv 4998 189067. MP_Wait 8898 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.035 40.807 40.807 qs_mol_dyn_low 1 2.0 0.003 0.003 40.562 40.570 qs_forces 11 3.9 0.002 0.003 40.498 40.499 qs_energies 11 4.9 0.004 0.007 38.580 38.583 scf_env_do_scf 11 5.9 0.001 0.001 32.654 32.654 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 24.816 24.818 velocity_verlet 10 3.0 0.002 0.002 22.694 22.700 dbcsr_multiply_generic 2286 12.5 0.098 0.100 17.881 18.075 qs_scf_new_mos 108 7.5 0.001 0.001 16.035 16.126 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.034 16.126 ot_scf_mini 108 9.5 0.002 0.002 14.954 15.050 multiply_cannon 2286 13.5 0.320 0.331 14.117 15.038 multiply_cannon_loop 2286 14.5 0.343 0.350 12.832 13.773 multiply_cannon_multrec 9144 15.5 3.579 5.456 8.857 9.134 ot_mini 108 10.5 0.001 0.001 8.935 9.048 init_scf_loop 11 6.9 0.000 0.000 7.811 7.814 qs_ot_get_derivative 108 11.5 0.001 0.001 6.895 6.994 rebuild_ks_matrix 119 8.3 0.000 0.000 6.756 6.897 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.756 6.896 prepare_preconditioner 11 7.9 0.000 0.000 6.867 6.881 make_preconditioner 11 8.9 0.000 0.000 6.867 6.881 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.426 6.757 dbcsr_mm_accdrv_process 12550 15.8 4.032 5.731 5.156 6.649 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.109 6.235 cp_fm_upper_to_full 72 14.2 3.214 4.598 3.214 4.598 init_scf_run 11 5.9 0.000 0.001 3.881 3.881 scf_env_initial_rho_setup 11 6.9 0.006 0.007 3.881 3.881 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.683 3.733 calculate_rho_elec 119 8.7 0.118 0.121 3.682 3.732 sum_up_and_integrate 119 10.3 0.065 0.066 3.612 3.618 mp_waitall_1 97218 16.6 2.625 3.590 2.625 3.590 integrate_v_rspace 119 11.3 0.004 0.004 3.547 3.554 qs_ot_get_p 119 10.4 0.001 0.001 3.321 3.455 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.596 2.997 dbcsr_complete_redistribute 329 12.2 0.290 0.294 2.009 2.822 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.185 2.528 apply_single 119 13.6 0.000 0.000 2.185 2.528 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.669 2.488 calculate_first_density_matrix 1 7.0 0.000 0.000 2.424 2.477 make_m2s 4572 13.5 0.038 0.038 2.240 2.424 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.399 2.344 calculate_dm_sparse 119 9.5 0.000 0.000 2.274 2.319 make_images 4572 14.5 0.353 0.381 2.121 2.305 mp_alltoall_i22 627 13.8 1.411 2.270 1.411 2.270 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.437 2.247 grid_integrate_task_list 119 12.3 2.004 2.034 2.004 2.034 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.022 2.024 ot_diis_step 108 11.5 0.014 0.015 2.016 2.017 pw_transfer 1439 11.6 0.066 0.067 2.014 2.017 jit_kernel_multiply 10 15.7 1.097 2.002 1.097 2.002 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.923 1.925 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 1.918 1.921 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.774 1.818 qs_energies_init_hamiltonians 11 5.9 0.023 0.026 1.774 1.775 density_rs2pw 119 9.7 0.003 0.003 1.743 1.767 mp_sum_l 7207 12.9 0.983 1.733 0.983 1.733 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.666 1.713 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.655 1.655 fft_wrap_pw1pw2_140 487 13.2 0.087 0.089 1.606 1.613 fft3d_ps 1201 14.6 0.567 0.584 1.551 1.554 grid_collocate_task_list 119 9.7 1.444 1.465 1.444 1.465 cp_fm_cholesky_invert 11 10.9 1.427 1.430 1.427 1.430 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.340 1.387 hybrid_alltoall_any 4725 16.4 0.087 0.148 1.112 1.377 rs_pw_transfer 974 11.9 0.009 0.009 1.323 1.346 wfi_extrapolate 11 7.9 0.001 0.001 1.341 1.341 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.331 1.332 cp_fm_diag_elpa_base 50 14.0 1.183 1.237 1.328 1.330 make_images_data 4572 15.5 0.042 0.046 1.061 1.296 mp_alltoall_d11v 2130 13.8 1.198 1.269 1.198 1.269 potential_pw2rs 119 12.3 0.014 0.015 1.163 1.166 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.136 1.156 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.095 1.108 multiply_cannon_sync_h2d 9144 15.5 1.041 1.044 1.041 1.044 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.983 1.016 acc_transpose_blocks 9144 15.5 0.038 0.039 0.971 0.994 qs_create_task_list 11 7.9 0.007 0.008 0.949 0.959 generate_qs_task_list 11 8.9 0.372 0.392 0.942 0.952 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.859 0.873 mp_alltoall_z22v 1201 16.6 0.847 0.869 0.847 0.869 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=40.807000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=707.363636, yerr=11.841934 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 500.178944E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 1009730. MP_ISendRecv 41800 9096. MP_Wait 58168 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.029 83.650 83.651 qs_mol_dyn_low 1 2.0 0.003 0.004 83.205 83.214 qs_forces 11 3.9 0.002 0.003 83.132 83.134 qs_energies 11 4.9 0.001 0.002 80.346 80.357 scf_env_do_scf 11 5.9 0.000 0.001 71.359 71.361 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 65.782 65.783 dbcsr_multiply_generic 2055 12.4 0.103 0.105 53.107 53.439 qs_scf_new_mos 99 7.5 0.000 0.001 48.858 48.957 qs_scf_loop_do_ot 99 8.5 0.000 0.001 48.857 48.957 ot_scf_mini 99 9.5 0.002 0.002 46.467 46.542 multiply_cannon 2055 13.4 0.177 0.183 43.239 44.271 multiply_cannon_loop 2055 14.4 1.556 1.596 42.274 43.235 velocity_verlet 10 3.0 0.001 0.002 41.692 41.693 ot_mini 99 10.5 0.001 0.001 28.522 28.594 qs_ot_get_derivative 99 11.5 0.001 0.001 21.645 21.745 multiply_cannon_multrec 49320 15.4 12.096 12.901 17.305 18.026 rebuild_ks_matrix 110 8.3 0.000 0.001 14.447 14.536 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.013 14.446 14.535 mp_waitall_1 241148 16.1 12.035 12.893 12.035 12.893 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.714 12.801 multiply_cannon_sync_h2d 49320 15.4 9.941 10.662 9.941 10.662 qs_ot_get_p 110 10.4 0.001 0.001 9.394 9.537 multiply_cannon_metrocomm3 49320 15.4 0.077 0.082 7.175 8.347 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.190 7.708 apply_single 110 13.6 0.000 0.001 7.190 7.708 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.035 7.593 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 7.267 7.324 sum_up_and_integrate 110 10.3 0.038 0.045 7.015 7.029 integrate_v_rspace 110 11.3 0.003 0.003 6.977 7.001 init_scf_run 11 5.9 0.000 0.001 6.988 6.989 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.988 6.988 ot_diis_step 99 11.5 0.005 0.006 6.564 6.564 mp_sum_l 6514 12.8 5.507 6.360 5.507 6.360 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.175 6.299 calculate_rho_elec 110 8.6 0.020 0.024 6.175 6.299 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.129 6.160 init_scf_loop 11 6.9 0.000 0.000 5.550 5.550 dbcsr_mm_accdrv_process 87628 16.1 2.072 2.185 5.089 5.400 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.115 5.115 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.634 4.635 cp_fm_redistribute_end 48 14.0 4.039 4.611 4.043 4.613 cp_fm_diag_elpa_base 48 14.0 0.563 4.516 0.567 4.542 make_m2s 4110 13.4 0.060 0.065 3.899 4.018 wfi_extrapolate 11 7.9 0.001 0.001 4.002 4.002 rs_pw_transfer 902 11.9 0.013 0.014 3.742 3.939 make_images 4110 14.4 0.178 0.192 3.805 3.928 calculate_dm_sparse 110 9.5 0.000 0.001 3.759 3.864 multiply_cannon_metrocomm1 49320 15.4 0.061 0.063 2.783 3.838 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.803 3.807 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.444 3.478 density_rs2pw 110 9.6 0.004 0.004 3.220 3.416 prepare_preconditioner 11 7.9 0.000 0.000 3.376 3.402 make_preconditioner 11 8.9 0.000 0.000 3.376 3.402 grid_integrate_task_list 110 12.3 3.239 3.364 3.239 3.364 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.280 3.344 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.152 3.208 calculate_first_density_matrix 1 7.0 0.000 0.000 2.905 2.908 pw_transfer 1331 11.6 0.054 0.065 2.842 2.901 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 2.755 2.815 jit_kernel_multiply 13 15.9 2.733 2.747 2.733 2.747 potential_pw2rs 110 12.3 0.006 0.007 2.614 2.635 mp_alltoall_d11v 2046 13.8 2.182 2.626 2.182 2.626 acc_transpose_blocks 49320 15.4 0.216 0.231 2.433 2.513 fft_wrap_pw1pw2_140 451 13.1 0.170 0.189 2.331 2.394 fft3d_ps 1111 14.6 0.742 0.828 2.291 2.341 grid_collocate_task_list 110 9.6 2.088 2.229 2.088 2.229 mp_waitany 14300 13.8 1.850 2.123 1.850 2.123 mp_sum_d 3883 11.9 1.480 2.007 1.480 2.007 make_images_data 4110 15.4 0.044 0.047 1.783 1.911 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.872 1.887 cp_fm_cholesky_invert 11 10.9 1.804 1.808 1.804 1.808 hybrid_alltoall_any 4261 16.3 0.082 0.478 1.561 1.806 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.651000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.000000, yerr=2.296242 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 585.089024E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 1941718. MP_ISendRecv 20680 26400. MP_Wait 32692 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.105 70.955 70.956 qs_mol_dyn_low 1 2.0 0.004 0.005 70.025 70.036 qs_forces 11 3.9 0.002 0.004 69.786 69.787 qs_energies 11 4.9 0.017 0.048 66.449 66.456 scf_env_do_scf 11 5.9 0.001 0.001 56.946 56.950 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.104 49.106 dbcsr_multiply_generic 2055 12.4 0.114 0.118 37.999 38.213 velocity_verlet 10 3.0 0.002 0.003 36.912 36.917 qs_scf_new_mos 99 7.5 0.001 0.001 32.955 33.102 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.954 33.101 multiply_cannon 2055 13.4 0.222 0.243 31.227 32.375 ot_scf_mini 99 9.5 0.003 0.004 31.250 31.386 multiply_cannon_loop 2055 14.4 0.932 0.962 29.992 31.127 ot_mini 99 10.5 0.001 0.001 18.359 18.479 multiply_cannon_multrec 24660 15.4 7.666 9.395 13.917 15.747 rebuild_ks_matrix 110 8.3 0.000 0.001 13.835 13.988 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.018 13.835 13.988 qs_ot_get_derivative 99 11.5 0.001 0.002 12.498 12.621 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.184 12.317 mp_waitall_1 186928 16.3 8.123 10.695 8.123 10.695 multiply_cannon_sync_h2d 24660 15.4 7.104 8.431 7.104 8.431 multiply_cannon_metrocomm3 24660 15.4 0.071 0.073 5.293 8.124 init_scf_loop 11 6.9 0.000 0.001 7.768 7.774 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.560 7.234 apply_single 110 13.6 0.000 0.001 6.559 7.234 init_scf_run 11 5.9 0.000 0.001 6.951 6.951 scf_env_initial_rho_setup 11 6.9 0.048 0.376 6.950 6.951 sum_up_and_integrate 110 10.3 0.053 0.061 6.666 6.675 integrate_v_rspace 110 11.3 0.002 0.003 6.612 6.625 dbcsr_mm_accdrv_process 52282 16.1 4.692 5.691 6.092 6.383 qs_ot_get_p 110 10.4 0.001 0.001 6.170 6.358 qs_rho_update_rho_low 110 7.6 0.001 0.004 6.226 6.282 calculate_rho_elec 110 8.6 0.039 0.047 6.225 6.281 ot_diis_step 99 11.5 0.010 0.012 5.807 5.808 prepare_preconditioner 11 7.9 0.000 0.000 5.683 5.701 make_preconditioner 11 8.9 0.000 0.001 5.683 5.701 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.822 5.572 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.269 5.421 make_m2s 4110 13.4 0.056 0.059 4.316 4.709 make_images 4110 14.4 0.403 0.447 4.213 4.603 qs_ot_p2m_diag 48 11.0 0.028 0.043 4.257 4.276 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.836 3.836 wfi_extrapolate 11 7.9 0.002 0.008 3.602 3.603 pw_transfer 1331 11.6 0.066 0.072 3.435 3.565 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.327 3.458 density_rs2pw 110 9.6 0.004 0.004 3.189 3.450 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.391 3.394 rs_pw_transfer 902 11.9 0.013 0.014 3.106 3.391 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.309 3.311 grid_integrate_task_list 110 12.3 3.151 3.304 3.151 3.304 cp_fm_redistribute_end 48 14.0 2.464 3.279 2.466 3.279 cp_fm_diag_elpa_base 48 14.0 0.776 3.151 0.809 3.243 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.152 3.212 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.013 3.061 calculate_dm_sparse 110 9.5 0.001 0.001 3.005 3.039 calculate_first_density_matrix 1 7.0 0.000 0.002 2.878 2.881 fft_wrap_pw1pw2_140 451 13.1 0.201 0.220 2.728 2.861 make_images_data 4110 15.4 0.047 0.051 2.449 2.851 fft3d_ps 1111 14.6 1.072 1.262 2.682 2.802 hybrid_alltoall_any 4261 16.3 0.103 0.448 2.145 2.749 cp_fm_cholesky_invert 11 10.9 2.649 2.655 2.649 2.655 mp_alltoall_d11v 2046 13.8 2.202 2.557 2.202 2.557 mp_sum_l 6514 12.8 1.898 2.515 1.898 2.515 potential_pw2rs 110 12.3 0.009 0.009 2.489 2.504 grid_collocate_task_list 110 9.6 2.118 2.275 2.118 2.275 jit_kernel_multiply 11 16.3 1.055 2.227 1.055 2.227 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.021 2.048 qs_energies_init_hamiltonians 11 5.9 0.008 0.017 1.950 1.950 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.814 1.832 multiply_cannon_metrocomm4 22605 15.4 0.073 0.077 0.763 1.810 mp_allgather_i34 2055 14.4 0.560 1.644 0.560 1.644 mp_irecv_dv 57340 16.2 0.641 1.606 0.641 1.606 mp_waitany 10164 13.8 1.288 1.594 1.288 1.594 acc_transpose_blocks 24660 15.4 0.106 0.109 1.530 1.555 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.539 1.552 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.374 1.481 dbcsr_complete_redistribute 325 12.2 0.251 0.305 1.205 1.473 mp_alltoall_z22v 1111 16.6 1.349 1.420 1.349 1.420 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.956000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=553.272727, yerr=6.634495 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 661.028864E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_ISendRecv 13640 27424. MP_Wait 32318 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.080 0.191 62.421 62.423 qs_mol_dyn_low 1 2.0 0.004 0.005 61.285 61.414 qs_forces 11 3.9 0.003 0.008 61.155 61.156 qs_energies 11 4.9 0.004 0.016 57.943 57.948 scf_env_do_scf 11 5.9 0.001 0.001 49.063 49.064 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 40.287 40.290 velocity_verlet 10 3.0 0.002 0.004 32.826 32.829 dbcsr_multiply_generic 2055 12.4 0.107 0.113 29.861 30.098 qs_scf_new_mos 99 7.5 0.001 0.001 25.688 25.791 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.687 25.790 ot_scf_mini 99 9.5 0.003 0.004 24.464 24.579 multiply_cannon 2055 13.4 0.214 0.227 23.398 24.513 multiply_cannon_loop 2055 14.4 0.620 0.635 22.228 23.197 ot_mini 99 10.5 0.001 0.001 14.234 14.351 rebuild_ks_matrix 110 8.3 0.000 0.000 12.292 12.407 qs_ks_build_kohn_sham_matrix 110 9.3 0.015 0.025 12.292 12.406 mp_waitall_1 146766 16.3 7.727 11.020 7.727 11.020 multiply_cannon_multrec 16440 15.4 3.876 4.801 10.145 10.967 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.834 10.939 qs_ot_get_derivative 99 11.5 0.001 0.002 9.721 9.840 init_scf_loop 11 6.9 0.000 0.001 8.741 8.748 multiply_cannon_metrocomm3 16440 15.4 0.042 0.043 4.810 7.985 prepare_preconditioner 11 7.9 0.000 0.000 6.923 6.938 make_preconditioner 11 8.9 0.000 0.000 6.923 6.938 sum_up_and_integrate 110 10.3 0.064 0.078 6.524 6.550 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.170 6.538 dbcsr_mm_accdrv_process 34862 16.1 5.041 5.821 6.124 6.482 integrate_v_rspace 110 11.3 0.003 0.004 6.460 6.477 init_scf_run 11 5.9 0.000 0.001 6.331 6.331 scf_env_initial_rho_setup 11 6.9 0.003 0.011 6.331 6.331 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.139 5.647 apply_single 110 13.6 0.000 0.000 5.139 5.647 qs_rho_update_rho_low 110 7.6 0.001 0.004 5.595 5.608 calculate_rho_elec 110 8.6 0.058 0.058 5.594 5.607 qs_ot_get_p 110 10.4 0.001 0.001 5.433 5.589 make_m2s 4110 13.4 0.050 0.051 4.159 4.504 ot_diis_step 99 11.5 0.011 0.013 4.483 4.483 make_images 4110 14.4 0.394 0.509 4.047 4.389 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.544 4.331 multiply_cannon_sync_h2d 16440 15.4 3.674 4.274 3.674 4.274 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.810 3.814 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.469 3.469 grid_integrate_task_list 110 12.3 3.192 3.358 3.192 3.358 calculate_first_density_matrix 1 7.0 0.001 0.006 3.267 3.269 pw_transfer 1331 11.6 0.065 0.072 3.242 3.254 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.135 3.149 calculate_dm_sparse 110 9.5 0.001 0.001 3.106 3.136 rs_pw_transfer 902 11.9 0.011 0.012 2.853 3.084 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.066 3.069 density_rs2pw 110 9.6 0.004 0.004 2.859 3.057 wfi_extrapolate 11 7.9 0.001 0.003 2.971 2.971 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.963 2.965 cp_fm_redistribute_end 48 14.0 1.844 2.940 1.846 2.941 cp_fm_diag_elpa_base 48 14.0 1.031 2.803 1.090 2.912 make_images_data 4110 15.4 0.044 0.048 2.393 2.842 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 2.683 2.732 hybrid_alltoall_any 4261 16.3 0.106 0.378 2.139 2.701 cp_fm_cholesky_invert 11 10.9 2.692 2.698 2.692 2.698 fft_wrap_pw1pw2_140 451 13.1 0.210 0.215 2.660 2.673 fft3d_ps 1111 14.6 1.066 1.082 2.448 2.462 mp_sum_l 6514 12.8 1.750 2.453 1.750 2.453 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.336 2.400 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.885 2.368 grid_collocate_task_list 110 9.6 2.177 2.317 2.177 2.317 potential_pw2rs 110 12.3 0.011 0.011 2.302 2.312 mp_irecv_dv 48980 15.7 0.815 2.242 0.815 2.242 mp_alltoall_d11v 2046 13.8 1.908 2.196 1.908 2.196 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 2.102 2.103 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.066 2.080 dbcsr_complete_redistribute 325 12.2 0.330 0.383 1.575 2.061 jit_kernel_multiply 8 16.3 0.702 1.913 0.702 1.913 cp_fm_upper_to_full 70 13.6 1.437 1.910 1.437 1.910 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.659 1.675 acc_transpose_blocks 16440 15.4 0.072 0.075 1.525 1.651 cp_fm_cholesky_decompose 22 10.9 1.584 1.607 1.584 1.607 copy_fm_to_dbcsr 174 11.2 0.001 0.002 1.096 1.569 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.358 1.503 mp_allgather_i34 2055 14.4 0.474 1.484 0.474 1.484 mp_waitany 17072 13.8 1.277 1.468 1.277 1.468 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.272 1.278 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.190 1.264 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=62.423000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.090909, yerr=10.774166 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 721.072128E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_ISendRecv 10120 43184. MP_Wait 25102 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.035 67.118 67.119 qs_mol_dyn_low 1 2.0 0.004 0.005 66.825 66.834 qs_forces 11 3.9 0.003 0.010 66.542 66.543 qs_energies 11 4.9 0.003 0.016 63.160 63.164 scf_env_do_scf 11 5.9 0.001 0.001 53.388 53.392 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 41.122 41.123 velocity_verlet 10 3.0 0.002 0.003 37.245 37.255 dbcsr_multiply_generic 2055 12.4 0.115 0.119 30.987 31.278 qs_scf_new_mos 99 7.5 0.001 0.001 26.645 26.753 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.644 26.752 ot_scf_mini 99 9.5 0.003 0.004 24.966 25.064 multiply_cannon 2055 13.4 0.247 0.264 23.474 24.337 multiply_cannon_loop 2055 14.4 0.882 0.901 22.048 22.661 multiply_cannon_multrec 24660 15.4 4.174 6.737 13.575 14.706 ot_mini 99 10.5 0.001 0.001 14.232 14.361 rebuild_ks_matrix 110 8.3 0.000 0.000 12.292 12.388 qs_ks_build_kohn_sham_matrix 110 9.3 0.030 0.151 12.291 12.388 init_scf_loop 11 6.9 0.001 0.005 12.225 12.228 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.911 10.999 dbcsr_mm_accdrv_process 52304 16.0 7.342 9.932 9.255 10.223 prepare_preconditioner 11 7.9 0.000 0.000 10.100 10.119 make_preconditioner 11 8.9 0.000 0.001 10.100 10.119 qs_ot_get_derivative 99 11.5 0.001 0.002 9.989 10.098 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.265 9.740 mp_waitall_1 126806 16.4 4.694 6.690 4.694 6.690 sum_up_and_integrate 110 10.3 0.100 0.325 6.349 6.576 init_scf_run 11 5.9 0.000 0.001 6.485 6.485 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.484 6.485 integrate_v_rspace 110 11.3 0.003 0.003 6.248 6.262 qs_ot_get_p 110 10.4 0.001 0.001 5.730 5.850 make_m2s 4110 13.4 0.059 0.060 5.496 5.812 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.756 5.763 calculate_rho_elec 110 8.6 0.077 0.081 5.755 5.762 make_images 4110 14.4 0.581 0.702 5.356 5.672 cp_fm_upper_to_full 70 13.8 3.368 4.787 3.368 4.787 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.100 4.210 apply_single 110 13.6 0.000 0.000 4.100 4.210 ot_diis_step 99 11.5 0.012 0.017 4.206 4.206 dbcsr_complete_redistribute 325 12.2 0.427 0.477 2.733 3.891 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.823 3.837 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.799 3.802 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 3.420 3.455 calculate_first_density_matrix 1 7.0 0.000 0.001 3.417 3.419 grid_integrate_task_list 110 12.3 3.277 3.396 3.277 3.396 copy_fm_to_dbcsr 174 11.2 0.001 0.002 2.209 3.353 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.299 3.349 multiply_cannon_sync_h2d 24660 15.4 3.182 3.337 3.182 3.337 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.323 3.323 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.527 3.310 pw_transfer 1331 11.6 0.065 0.074 3.266 3.291 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.159 3.187 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.106 3.171 make_images_data 4110 15.4 0.046 0.050 2.772 3.116 hybrid_alltoall_any 4261 16.3 0.121 0.461 2.452 3.004 density_rs2pw 110 9.6 0.004 0.004 2.869 3.003 calculate_dm_sparse 110 9.5 0.001 0.001 2.967 2.992 wfi_extrapolate 11 7.9 0.001 0.001 2.984 2.984 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.826 2.950 mp_alltoall_i22 605 13.7 1.688 2.888 1.688 2.888 jit_kernel_multiply 10 15.7 1.585 2.806 1.585 2.806 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.777 2.779 cp_fm_redistribute_end 48 14.0 1.384 2.750 1.386 2.751 cp_fm_cholesky_invert 11 10.9 2.723 2.731 2.723 2.731 cp_fm_diag_elpa_base 48 14.0 1.281 2.613 1.362 2.727 rs_pw_transfer 902 11.9 0.011 0.011 2.561 2.717 fft_wrap_pw1pw2_140 451 13.1 0.202 0.212 2.687 2.716 qs_energies_init_hamiltonians 11 5.9 0.134 0.419 2.705 2.707 fft3d_ps 1111 14.6 1.063 1.090 2.461 2.482 grid_collocate_task_list 110 9.6 2.226 2.352 2.226 2.352 potential_pw2rs 110 12.3 0.012 0.013 2.096 2.104 mp_alltoall_d11v 2046 13.8 1.877 2.043 1.877 2.043 mp_sum_l 6514 12.8 1.331 1.970 1.331 1.970 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.748 1.786 cp_fm_cholesky_decompose 22 10.9 1.706 1.751 1.706 1.751 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.605 1.704 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.630 1.642 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.293 1.602 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.525 1.541 acc_transpose_blocks 24660 15.4 0.104 0.106 1.501 1.530 multiply_cannon_metrocomm4 20550 15.4 0.056 0.059 0.849 1.469 mp_sum_d 3881 11.9 0.924 1.441 0.924 1.441 mp_irecv_dv 62702 16.1 0.752 1.394 0.752 1.394 mp_allgather_i34 2055 14.4 0.520 1.353 0.520 1.353 mp_waitany 13376 13.8 1.138 1.351 1.138 1.351 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.119000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=684.818182, yerr=7.468944 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/13/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 807.299199E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 1438408 0.0% 0.0% 100.0% average stack size 0.0 0.0 684.2 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 829.911040E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_ISendRecv 6600 54848. MP_Wait 17226 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.031 59.272 59.272 qs_mol_dyn_low 1 2.0 0.003 0.003 59.009 59.020 qs_forces 11 3.9 0.002 0.002 58.909 58.912 qs_energies 11 4.9 0.001 0.002 55.292 55.297 scf_env_do_scf 11 5.9 0.000 0.001 43.540 43.540 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 35.361 35.361 velocity_verlet 10 3.0 0.002 0.002 31.363 31.368 dbcsr_multiply_generic 2055 12.4 0.103 0.105 26.569 26.742 multiply_cannon 2055 13.4 0.248 0.260 20.904 22.321 qs_scf_new_mos 99 7.5 0.001 0.001 21.002 21.055 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.001 21.054 multiply_cannon_loop 2055 14.4 0.323 0.335 19.501 19.825 ot_scf_mini 99 9.5 0.002 0.002 19.745 19.777 multiply_cannon_multrec 8220 15.4 3.221 4.468 10.733 11.754 rebuild_ks_matrix 110 8.3 0.000 0.000 11.568 11.595 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.568 11.594 ot_mini 99 10.5 0.001 0.001 10.579 10.605 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.333 10.357 init_scf_run 11 5.9 0.000 0.001 8.611 8.611 scf_env_initial_rho_setup 11 6.9 0.001 0.001 8.611 8.611 dbcsr_mm_accdrv_process 17442 15.9 3.057 6.178 7.381 8.299 init_scf_loop 11 6.9 0.000 0.000 8.132 8.133 mp_waitall_1 106626 16.5 6.416 8.037 6.416 8.037 qs_ot_get_derivative 99 11.5 0.001 0.001 6.758 6.788 prepare_preconditioner 11 7.9 0.000 0.000 6.504 6.510 make_preconditioner 11 8.9 0.000 0.000 6.504 6.510 sum_up_and_integrate 110 10.3 0.079 0.081 6.261 6.274 integrate_v_rspace 110 11.3 0.003 0.003 6.181 6.194 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.045 6.121 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 6.046 6.047 calculate_first_density_matrix 1 7.0 0.000 0.000 5.780 5.783 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.740 5.760 calculate_rho_elec 110 8.6 0.115 0.115 5.739 5.759 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.636 5.651 qs_ot_get_p 110 10.4 0.001 0.001 5.203 5.244 make_m2s 4110 13.4 0.038 0.040 4.343 4.661 jit_kernel_multiply 7 15.7 4.021 4.554 4.021 4.554 make_images 4110 14.4 0.638 0.698 4.216 4.533 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 3.010 4.402 ot_diis_step 99 11.5 0.012 0.012 3.792 3.793 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.775 3.779 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.706 3.725 apply_single 110 13.6 0.000 0.000 3.706 3.725 grid_integrate_task_list 110 12.3 3.360 3.469 3.360 3.469 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.444 3.445 pw_transfer 1331 11.6 0.065 0.070 3.284 3.295 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.177 3.192 cp_fm_cholesky_invert 11 10.9 3.144 3.149 3.144 3.149 multiply_cannon_sync_h2d 8220 15.4 2.912 3.061 2.912 3.061 density_rs2pw 110 9.6 0.004 0.004 2.761 3.004 make_images_data 4110 15.4 0.038 0.043 2.479 2.969 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.872 2.873 cp_fm_redistribute_end 48 14.0 0.729 2.838 0.734 2.839 hybrid_alltoall_any 4261 16.3 0.200 0.863 2.369 2.823 cp_fm_diag_elpa_base 48 14.0 1.917 2.634 2.095 2.799 wfi_extrapolate 11 7.9 0.001 0.001 2.744 2.744 fft_wrap_pw1pw2_140 451 13.1 0.212 0.216 2.681 2.699 qs_energies_init_hamiltonians 11 5.9 0.017 0.033 2.673 2.691 rs_pw_transfer 902 11.9 0.010 0.011 2.333 2.608 calculate_dm_sparse 110 9.5 0.001 0.001 2.500 2.544 fft3d_ps 1111 14.6 1.114 1.141 2.433 2.442 grid_collocate_task_list 110 9.6 2.322 2.434 2.322 2.434 potential_pw2rs 110 12.3 0.015 0.016 2.000 2.004 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.776 1.998 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.956 1.971 cp_fm_cholesky_decompose 22 10.9 1.864 1.884 1.864 1.884 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.854 1.873 mp_alltoall_d11v 2046 13.8 1.743 1.858 1.743 1.858 mp_allgather_i34 2055 14.4 0.583 1.716 0.583 1.716 dbcsr_complete_redistribute 325 12.2 0.574 0.642 1.572 1.673 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.503 1.614 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.568 1.572 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.445 1.458 mp_waitany 9240 13.8 1.107 1.386 1.107 1.386 qs_create_task_list 11 7.9 0.002 0.004 1.235 1.329 generate_qs_task_list 11 8.9 0.378 0.445 1.233 1.328 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.866 1.254 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.164 1.191 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=59.272000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=784.818182, yerr=9.103627 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/14/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 1.612391E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 1464624 0.0% 0.0% 100.0% average stack size 0.0 0.0 672.0 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 1.367953E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_ISendRecv 4620 360267. MP_Wait 7524 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.059 86.855 86.856 qs_mol_dyn_low 1 2.0 0.003 0.003 85.993 86.004 qs_forces 11 3.9 0.002 0.002 85.913 85.914 qs_energies 11 4.9 0.001 0.001 81.885 81.887 scf_env_do_scf 11 5.9 0.001 0.001 71.097 71.097 velocity_verlet 10 3.0 0.002 0.002 54.639 54.647 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 42.694 42.697 dbcsr_multiply_generic 2055 12.4 0.118 0.120 29.395 29.455 init_scf_loop 11 6.9 0.000 0.000 28.321 28.324 prepare_preconditioner 11 7.9 0.000 0.000 26.360 26.366 make_preconditioner 11 8.9 0.000 0.000 26.360 26.366 qs_scf_new_mos 99 7.5 0.001 0.001 26.178 26.214 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.178 26.213 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.487 25.824 ot_scf_mini 99 9.5 0.002 0.002 24.436 24.455 multiply_cannon 2055 13.4 0.346 0.378 22.405 23.271 multiply_cannon_loop 2055 14.4 0.341 0.344 20.637 21.001 cp_fm_upper_to_full 70 14.2 12.721 18.308 12.721 18.308 ot_mini 99 10.5 0.001 0.001 13.671 13.691 rebuild_ks_matrix 110 8.3 0.000 0.001 13.243 13.268 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.243 13.267 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.043 12.066 dbcsr_complete_redistribute 325 12.2 1.037 1.075 7.492 10.668 multiply_cannon_multrec 8220 15.4 4.404 4.603 10.173 10.310 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.441 9.622 qs_ot_get_derivative 99 11.5 0.001 0.001 9.190 9.210 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.857 9.001 mp_waitall_1 87304 16.6 7.841 8.732 7.841 8.732 mp_alltoall_i22 605 13.7 5.465 8.653 5.465 8.653 sum_up_and_integrate 110 10.3 0.150 0.151 6.715 6.729 integrate_v_rspace 110 11.3 0.004 0.004 6.565 6.578 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.533 6.570 calculate_rho_elec 110 8.6 0.227 0.227 6.532 6.570 init_scf_run 11 5.9 0.000 0.001 6.478 6.480 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.478 6.479 dbcsr_mm_accdrv_process 11614 15.7 3.301 3.794 5.626 5.886 make_m2s 4110 13.4 0.043 0.044 5.309 5.802 qs_ot_get_p 110 10.4 0.001 0.001 5.632 5.655 make_images 4110 14.4 0.884 0.935 5.122 5.615 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.773 5.304 cp_fm_cholesky_invert 11 10.9 5.219 5.224 5.219 5.224 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.613 5.082 apply_single 110 13.6 0.000 0.000 4.612 5.082 ot_diis_step 99 11.5 0.015 0.016 4.467 4.467 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.065 4.073 multiply_cannon_sync_h2d 8220 15.4 3.950 3.952 3.950 3.952 pw_transfer 1331 11.6 0.074 0.075 3.891 3.896 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.775 3.779 calculate_dm_sparse 110 9.5 0.001 0.001 3.718 3.734 grid_integrate_task_list 110 12.3 3.655 3.707 3.655 3.707 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.703 3.705 hybrid_alltoall_any 4261 16.3 0.256 0.552 2.902 3.648 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.626 3.627 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.094 3.550 make_images_data 4110 15.4 0.042 0.045 2.894 3.537 fft_wrap_pw1pw2_140 451 13.1 0.215 0.218 3.262 3.268 wfi_extrapolate 11 7.9 0.001 0.001 3.218 3.218 calculate_first_density_matrix 1 7.0 0.000 0.000 3.148 3.152 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.126 3.128 density_rs2pw 110 9.6 0.004 0.004 3.031 3.050 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.035 3.036 cp_fm_diag_elpa_base 48 14.0 2.502 2.702 3.033 3.033 fft3d_ps 1111 14.6 1.264 1.269 2.982 2.987 grid_collocate_task_list 110 9.6 2.626 2.650 2.626 2.650 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.616 2.626 jit_kernel_multiply 10 15.4 2.126 2.424 2.126 2.424 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.366 2.379 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.242 2.298 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.274 2.290 rs_pw_transfer 902 11.9 0.011 0.011 2.176 2.251 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.090 2.184 mp_alltoall_d11v 2046 13.8 2.084 2.155 2.084 2.155 potential_pw2rs 110 12.3 0.021 0.021 2.137 2.139 cp_fm_cholesky_decompose 22 10.9 2.013 2.030 2.013 2.030 qs_create_task_list 11 7.9 0.005 0.006 1.915 1.958 generate_qs_task_list 11 8.9 0.740 0.794 1.909 1.953 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.782 1.819 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.807 1.808 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.856000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1222.545455, yerr=60.316383 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 626.581504E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 56760. MP_Allreduce 11253 785. MP_Sync 170 MP_Alltoall 2226 1581565. MP_ISendRecv 48640 18752. MP_Wait 66796 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.047 201.858 201.859 qs_mol_dyn_low 1 2.0 0.003 0.004 201.447 201.461 qs_forces 11 3.9 0.003 0.005 200.882 200.885 qs_energies 11 4.9 0.001 0.002 195.437 195.452 scf_env_do_scf 11 5.9 0.001 0.002 179.032 179.036 scf_env_do_scf_inner_loop 117 6.6 0.002 0.008 158.681 158.684 dbcsr_multiply_generic 2507 12.6 0.171 0.175 123.816 124.434 velocity_verlet 10 3.0 0.001 0.002 121.370 121.376 qs_scf_new_mos 117 7.6 0.001 0.001 120.632 120.918 qs_scf_loop_do_ot 117 8.6 0.001 0.001 120.632 120.917 ot_scf_mini 117 9.6 0.003 0.003 114.017 114.280 multiply_cannon 2507 13.6 0.241 0.251 100.611 102.685 multiply_cannon_loop 2507 14.6 2.100 2.177 98.518 100.146 ot_mini 117 10.6 0.001 0.001 65.407 65.679 multiply_cannon_multrec 60168 15.6 33.051 34.892 41.455 43.021 qs_ot_get_derivative 117 11.6 0.001 0.001 40.583 40.861 rebuild_ks_matrix 128 8.3 0.001 0.001 32.831 33.154 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.017 32.830 33.154 mp_waitall_1 291448 16.2 28.340 31.440 28.340 31.440 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.536 29.845 multiply_cannon_sync_h2d 60168 15.6 27.303 29.681 27.303 29.681 qs_ot_get_p 128 10.4 0.001 0.001 26.713 27.037 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.222 24.799 apply_single 128 13.6 0.001 0.001 24.222 24.799 ot_diis_step 117 11.6 0.007 0.008 24.439 24.440 init_scf_loop 11 6.9 0.000 0.000 20.278 20.280 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.044 20.152 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.013 19.233 multiply_cannon_metrocomm3 60168 15.6 0.114 0.118 15.593 18.672 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.528 17.529 prepare_preconditioner 11 7.9 0.000 0.000 15.775 15.811 make_preconditioner 11 8.9 0.000 0.000 15.775 15.811 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.020 15.212 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.732 14.737 cp_fm_redistribute_end 83 14.4 11.638 14.664 11.650 14.666 cp_fm_diag_elpa_base 83 14.4 2.974 14.368 3.006 14.489 make_m2s 5014 13.6 0.102 0.109 13.565 13.937 make_images 5014 14.6 0.410 0.429 13.392 13.771 sum_up_and_integrate 128 10.3 0.091 0.109 13.660 13.676 integrate_v_rspace 128 11.3 0.003 0.004 13.569 13.589 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.272 12.368 calculate_rho_elec 128 8.7 0.045 0.063 12.271 12.368 init_scf_run 11 5.9 0.000 0.001 12.268 12.269 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.268 12.268 mp_sum_l 7870 13.0 8.401 9.602 8.401 9.602 cp_fm_cholesky_invert 11 10.9 8.988 8.996 8.988 8.996 wfi_extrapolate 11 7.9 0.001 0.001 8.902 8.902 dbcsr_mm_accdrv_process 124484 16.2 3.275 3.453 7.971 8.588 calculate_dm_sparse 128 9.5 0.001 0.001 8.400 8.513 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.866 8.019 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.864 7.941 multiply_cannon_metrocomm1 60168 15.6 0.088 0.093 6.139 7.610 make_images_data 5014 15.6 0.073 0.079 6.602 7.539 grid_integrate_task_list 128 12.3 7.092 7.534 7.092 7.534 hybrid_alltoall_any 5200 16.5 0.292 2.256 5.789 6.923 density_rs2pw 128 9.7 0.006 0.007 6.167 6.711 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.688 6.696 pw_transfer 1547 11.6 0.076 0.107 6.354 6.620 fft_wrap_pw1pw2 1291 12.7 0.010 0.014 6.149 6.385 rs_pw_transfer 1046 11.9 0.018 0.020 5.591 6.130 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.873 5.972 fft_wrap_pw1pw2_140 523 13.2 0.447 0.510 5.310 5.479 fft3d_ps 1291 14.7 2.087 2.569 4.936 5.118 grid_collocate_task_list 128 9.7 4.723 5.061 4.723 5.061 mp_alltoall_d11v 2415 14.1 3.881 5.056 3.881 5.056 cp_fm_cholesky_decompose 22 10.9 4.642 4.659 4.642 4.659 mp_sum_d 4464 12.1 3.590 4.551 3.590 4.551 potential_pw2rs 128 12.3 0.010 0.011 4.257 4.279 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=201.859000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.090909, yerr=6.801798 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 833.089536E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453843137296 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4079 57283. MP_Allreduce 11237 947. MP_Sync 168 MP_Alltoall 1955 6686093. MP_ISendRecv 23876 47072. MP_Wait 37656 MP_comm_split 82 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.046 191.940 191.942 qs_mol_dyn_low 1 2.0 0.008 0.036 190.865 190.901 qs_forces 11 3.9 0.011 0.050 190.670 190.681 qs_energies 11 4.9 0.003 0.013 183.903 183.916 scf_env_do_scf 11 5.9 0.001 0.010 166.606 166.616 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 132.776 132.779 velocity_verlet 10 3.0 0.001 0.002 119.363 119.367 dbcsr_multiply_generic 2485 12.5 0.181 0.186 98.022 99.291 qs_scf_new_mos 116 7.6 0.001 0.001 94.323 94.845 qs_scf_loop_do_ot 116 8.6 0.001 0.001 94.322 94.844 ot_scf_mini 116 9.6 0.004 0.005 89.550 90.114 multiply_cannon 2485 13.5 0.475 0.530 77.721 82.344 multiply_cannon_loop 2485 14.5 1.244 1.279 74.332 76.694 ot_mini 116 10.6 0.001 0.002 49.612 50.112 mp_waitall_1 224796 16.4 25.254 38.757 25.254 38.757 multiply_cannon_multrec 29820 15.5 21.733 25.967 31.690 36.243 init_scf_loop 11 6.9 0.000 0.001 33.731 33.733 rebuild_ks_matrix 127 8.3 0.001 0.001 32.736 33.254 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.025 32.735 33.253 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.478 29.953 prepare_preconditioner 11 7.9 0.000 0.000 29.278 29.336 make_preconditioner 11 8.9 0.000 0.000 29.278 29.336 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.926 28.466 qs_ot_get_derivative 116 11.6 0.046 0.276 27.788 28.350 multiply_cannon_metrocomm3 29820 15.5 0.096 0.102 15.797 27.765 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.788 22.987 apply_single 127 13.6 0.001 0.001 21.788 22.987 qs_ot_get_p 127 10.4 0.001 0.001 21.701 22.329 ot_diis_step 116 11.6 0.021 0.040 21.650 21.667 multiply_cannon_sync_h2d 29820 15.5 19.044 20.948 19.044 20.948 cp_fm_cholesky_invert 11 10.9 17.105 17.118 17.105 17.118 qs_ot_p2m_diag 82 11.4 0.185 0.214 16.852 16.892 make_m2s 4970 13.5 0.090 0.094 14.474 16.126 make_images 4970 14.5 1.159 1.357 14.276 15.929 cp_dbcsr_syevd 82 12.4 0.005 0.005 15.695 15.696 sum_up_and_integrate 127 10.3 0.116 0.133 14.310 14.346 integrate_v_rspace 127 11.3 0.004 0.005 14.194 14.234 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.926 12.966 calculate_rho_elec 127 8.7 0.087 0.104 12.926 12.965 cp_fm_diag_elpa 82 13.4 0.000 0.001 12.561 12.572 cp_fm_redistribute_end 82 14.4 7.329 12.485 7.343 12.486 init_scf_run 11 5.9 0.000 0.001 12.423 12.424 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.422 12.424 cp_fm_diag_elpa_base 82 14.4 4.907 12.011 5.122 12.367 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.912 11.348 make_images_data 4970 15.5 0.067 0.073 8.823 10.710 multiply_cannon_metrocomm4 27335 15.5 0.095 0.112 3.744 10.376 hybrid_alltoall_any 5155 16.4 0.341 1.500 7.411 10.014 dbcsr_mm_accdrv_process 61726 16.2 4.740 5.641 9.428 10.011 mp_irecv_dv 68888 16.3 3.554 10.006 3.554 10.006 wfi_extrapolate 11 7.9 0.002 0.009 8.449 8.449 pw_transfer 1535 11.6 0.085 0.100 7.738 7.806 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.515 7.580 grid_integrate_task_list 127 12.3 7.091 7.496 7.091 7.496 calculate_dm_sparse 127 9.5 0.001 0.001 7.200 7.362 cp_fm_cholesky_decompose 22 10.9 7.189 7.273 7.189 7.273 density_rs2pw 127 9.7 0.006 0.006 6.799 7.156 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.236 7.023 fft_wrap_pw1pw2_140 519 13.2 0.465 0.512 6.519 6.592 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.216 6.228 rs_pw_transfer 1038 11.9 0.015 0.017 5.707 6.148 fft3d_ps 1281 14.7 2.752 2.908 5.897 5.926 mp_sum_l 7804 13.0 4.074 5.672 4.074 5.672 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.342 5.496 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.307 5.361 grid_collocate_task_list 127 9.7 4.868 5.247 4.868 5.247 mp_allgather_i34 2485 14.5 1.965 5.189 1.965 5.189 potential_pw2rs 127 12.3 0.015 0.017 4.806 4.823 mp_alltoall_d11v 2401 14.1 4.164 4.768 4.164 4.768 dbcsr_complete_redistribute 393 12.7 0.771 0.880 3.273 4.122 mp_sum_d 4443 12.1 2.737 4.052 2.737 4.052 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=191.942000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=789.909091, yerr=4.813892 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 974.282752E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57335. MP_Allreduce 11226 986. MP_Sync 170 MP_Alltoall 1712 9388896. MP_ISendRecv 15872 75008. MP_Wait 29756 MP_comm_split 83 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.069 176.650 176.651 qs_mol_dyn_low 1 2.0 0.003 0.004 176.200 176.213 qs_forces 11 3.9 0.003 0.003 176.103 176.106 qs_energies 11 4.9 0.001 0.002 169.634 169.644 scf_env_do_scf 11 5.9 0.001 0.002 153.115 153.117 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 117.690 117.690 velocity_verlet 10 3.0 0.002 0.003 112.758 112.760 dbcsr_multiply_generic 2507 12.6 0.179 0.183 82.258 83.428 qs_scf_new_mos 117 7.6 0.001 0.001 81.585 81.866 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.584 81.865 ot_scf_mini 117 9.6 0.003 0.004 77.460 77.826 multiply_cannon 2507 13.6 0.503 0.520 62.661 66.416 multiply_cannon_loop 2507 14.6 0.861 0.887 59.258 61.454 ot_mini 117 10.6 0.001 0.001 42.541 42.915 init_scf_loop 11 6.9 0.000 0.000 35.324 35.325 mp_waitall_1 178456 16.5 24.478 33.637 24.478 33.637 prepare_preconditioner 11 7.9 0.000 0.000 31.373 31.425 make_preconditioner 11 8.9 0.000 0.000 31.373 31.425 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.933 30.381 rebuild_ks_matrix 128 8.3 0.001 0.001 29.696 30.176 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.017 29.695 30.175 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.816 27.244 multiply_cannon_multrec 20056 15.6 13.750 16.880 22.771 25.856 multiply_cannon_metrocomm3 20056 15.6 0.059 0.062 14.683 23.661 qs_ot_get_derivative 117 11.6 0.001 0.002 22.728 23.084 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.898 20.873 apply_single 128 13.6 0.001 0.001 19.898 20.873 qs_ot_get_p 128 10.4 0.001 0.001 20.333 20.777 ot_diis_step 117 11.6 0.018 0.018 19.707 19.708 qs_ot_p2m_diag 83 11.4 0.265 0.272 16.017 16.026 multiply_cannon_sync_h2d 20056 15.6 14.351 15.789 14.351 15.789 make_m2s 5014 13.6 0.082 0.086 14.413 15.547 make_images 5014 14.6 1.189 1.299 14.185 15.318 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.002 15.004 cp_fm_cholesky_invert 11 10.9 14.608 14.617 14.608 14.617 sum_up_and_integrate 128 10.3 0.134 0.147 13.932 13.954 integrate_v_rspace 128 11.3 0.004 0.004 13.797 13.822 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.847 12.898 calculate_rho_elec 128 8.7 0.132 0.147 12.846 12.898 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.970 11.970 cp_fm_redistribute_end 83 14.4 4.508 11.906 4.523 11.910 cp_fm_diag_elpa_base 83 14.4 6.950 11.307 7.364 11.798 init_scf_run 11 5.9 0.000 0.001 11.469 11.469 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.469 11.469 make_images_data 5014 15.6 0.062 0.068 8.806 10.305 hybrid_alltoall_any 5200 16.5 0.433 1.976 7.698 9.481 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.970 9.254 multiply_cannon_metrocomm4 17549 15.6 0.062 0.073 3.327 8.938 mp_irecv_dv 50230 16.2 3.205 8.694 3.205 8.694 dbcsr_mm_accdrv_process 41502 16.2 4.682 5.562 8.477 8.670 grid_integrate_task_list 128 12.3 7.298 7.692 7.298 7.692 cp_fm_cholesky_decompose 22 10.9 7.542 7.614 7.542 7.614 pw_transfer 1547 11.6 0.085 0.106 7.476 7.579 cp_fm_upper_to_full 105 14.5 5.762 7.455 5.762 7.455 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.253 7.362 wfi_extrapolate 11 7.9 0.001 0.001 7.256 7.256 density_rs2pw 128 9.7 0.006 0.006 6.307 6.742 fft_wrap_pw1pw2_140 523 13.2 0.479 0.534 6.372 6.488 dbcsr_complete_redistribute 395 12.7 1.168 1.194 4.584 6.379 calculate_dm_sparse 128 9.5 0.001 0.001 6.285 6.377 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.764 5.770 grid_collocate_task_list 128 9.7 5.090 5.605 5.090 5.605 fft3d_ps 1291 14.7 2.692 2.902 5.518 5.579 rs_pw_transfer 1046 11.9 0.014 0.015 4.943 5.430 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.398 5.187 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.464 5.173 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.874 4.990 mp_sum_l 7870 13.0 3.359 4.654 3.359 4.654 mp_alltoall_d11v 2415 14.1 4.112 4.591 4.112 4.591 mp_allgather_i34 2507 14.6 1.912 4.589 1.912 4.589 potential_pw2rs 128 12.3 0.020 0.022 4.365 4.376 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.421 4.153 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.022 4.049 calculate_first_density_matrix 1 7.0 0.000 0.000 4.043 4.045 mp_alltoall_i22 716 14.1 2.013 3.919 2.013 3.919 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.724 3.725 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=176.651000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=913.909091, yerr=20.685764 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420243808256 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528908111872 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514772E+12 0.0% 0.0% 100.0% flops max/rank 4.353792E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755942624 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.181815E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379572320 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4103 56948. MP_Allreduce 11296 1066. MP_Sync 170 MP_Alltoall 1712 12503107. MP_ISendRecv 11776 75008. MP_Wait 28330 MP_comm_split 83 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.052 0.122 189.272 189.273 qs_mol_dyn_low 1 2.0 0.007 0.031 188.626 188.690 qs_forces 11 3.9 0.023 0.096 188.419 188.456 qs_energies 11 4.9 0.005 0.019 181.360 181.376 scf_env_do_scf 11 5.9 0.001 0.002 163.886 163.896 velocity_verlet 10 3.0 0.004 0.020 124.599 124.605 scf_env_do_scf_inner_loop 117 6.6 0.004 0.008 116.830 116.832 qs_scf_new_mos 117 7.6 0.001 0.001 81.506 81.783 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.505 81.782 dbcsr_multiply_generic 2507 12.6 0.189 0.194 80.002 80.757 ot_scf_mini 117 9.6 0.004 0.010 76.997 77.300 multiply_cannon 2507 13.6 0.560 0.587 55.326 57.855 multiply_cannon_loop 2507 14.6 1.180 1.217 51.432 53.293 init_scf_loop 11 6.9 0.000 0.001 46.930 46.933 prepare_preconditioner 11 7.9 0.000 0.000 42.902 42.923 make_preconditioner 11 8.9 0.000 0.001 42.902 42.923 ot_mini 117 10.6 0.001 0.002 42.386 42.694 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.430 41.490 multiply_cannon_multrec 30084 15.6 14.219 19.744 26.427 31.235 rebuild_ks_matrix 128 8.3 0.001 0.001 28.777 29.078 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.023 28.776 29.078 mp_waitall_1 153770 16.5 17.654 27.616 17.654 27.616 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.019 26.279 qs_ot_get_derivative 117 11.6 0.001 0.002 22.753 23.059 make_m2s 5014 13.6 0.097 0.101 20.223 21.270 qs_ot_get_p 128 10.4 0.001 0.001 20.798 21.160 make_images 5014 14.6 1.961 2.313 19.920 20.971 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 18.993 19.544 apply_single 128 13.6 0.001 0.001 18.992 19.544 ot_diis_step 117 11.6 0.021 0.030 19.499 19.504 qs_ot_p2m_diag 83 11.4 0.343 0.390 16.640 16.690 cp_fm_upper_to_full 105 14.7 11.182 16.584 11.182 16.584 cp_fm_cholesky_invert 11 10.9 16.433 16.442 16.433 16.442 multiply_cannon_metrocomm3 30084 15.6 0.047 0.049 6.430 15.719 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.341 15.342 sum_up_and_integrate 128 10.3 0.141 0.155 13.931 13.954 integrate_v_rspace 128 11.3 0.004 0.004 13.789 13.818 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.073 13.126 calculate_rho_elec 128 8.7 0.175 0.190 13.072 13.125 dbcsr_complete_redistribute 395 12.7 1.514 1.661 9.174 12.955 multiply_cannon_sync_h2d 30084 15.6 11.696 12.555 11.696 12.555 make_images_data 5014 15.6 0.066 0.069 10.815 12.499 dbcsr_mm_accdrv_process 62264 16.2 7.633 8.615 11.780 12.299 cp_fm_diag_elpa 83 13.4 0.000 0.000 12.157 12.160 cp_fm_redistribute_end 83 14.4 2.088 12.063 2.103 12.069 cp_fm_diag_elpa_base 83 14.4 9.339 11.436 9.930 11.928 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.748 11.533 init_scf_run 11 5.9 0.000 0.001 11.462 11.464 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.462 11.463 hybrid_alltoall_any 5200 16.5 0.527 2.215 9.688 11.375 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.449 10.150 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.434 9.663 mp_alltoall_i22 716 14.1 5.650 9.522 5.650 9.522 cp_fm_cholesky_decompose 22 10.9 7.764 7.859 7.764 7.859 grid_integrate_task_list 128 12.3 7.506 7.857 7.506 7.857 wfi_extrapolate 11 7.9 0.005 0.038 7.675 7.675 pw_transfer 1547 11.6 0.085 0.099 7.592 7.675 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.367 7.456 multiply_cannon_metrocomm4 25070 15.6 0.075 0.084 2.769 6.976 mp_irecv_dv 76098 16.2 2.626 6.715 2.626 6.715 density_rs2pw 128 9.7 0.005 0.006 6.243 6.691 fft_wrap_pw1pw2_140 523 13.2 0.481 0.488 6.481 6.582 calculate_dm_sparse 128 9.5 0.001 0.001 6.487 6.562 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.893 5.951 fft3d_ps 1291 14.7 2.796 2.877 5.596 5.668 grid_collocate_task_list 128 9.7 5.204 5.598 5.204 5.598 rs_pw_transfer 1046 11.9 0.014 0.015 4.758 5.209 mp_alltoall_d11v 2415 14.1 4.745 5.192 4.745 5.192 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 4.756 4.840 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 4.479 4.491 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.354 4.445 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.234 4.293 potential_pw2rs 128 12.3 0.023 0.023 4.220 4.234 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=189.273000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1095.545455, yerr=21.722898 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 5.909858E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 1976192 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.2 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.498886E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242592 MPI messages size (bytes): total size 1.341705E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.530704E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 114816 60196651008 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726591885696 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9003 51. MP_Alltoall 9708 795382. MP_ISend 40388 2101966. MP_IRecv 40388 2101070. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57898. MP_Allreduce 11057 1172. MP_Sync 87 MP_Alltoall 1724 18848418. MP_ISendRecv 7740 122880. MP_Wait 20114 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.045 174.348 174.349 qs_mol_dyn_low 1 2.0 0.003 0.003 173.851 173.863 qs_forces 11 3.9 0.003 0.003 173.618 173.624 qs_energies 11 4.9 0.002 0.002 166.321 166.330 scf_env_do_scf 11 5.9 0.001 0.009 148.540 148.553 velocity_verlet 10 3.0 0.012 0.014 113.601 113.614 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 112.469 112.471 dbcsr_multiply_generic 2527 12.6 0.179 0.186 76.355 76.892 qs_scf_new_mos 118 7.6 0.001 0.001 76.356 76.499 qs_scf_loop_do_ot 118 8.6 0.001 0.001 76.355 76.498 ot_scf_mini 118 9.6 0.004 0.004 71.847 72.044 multiply_cannon 2527 13.6 0.588 0.617 55.939 60.294 multiply_cannon_loop 2527 14.6 0.454 0.470 51.196 52.356 ot_mini 118 10.6 0.001 0.001 40.557 40.755 init_scf_loop 11 6.9 0.000 0.000 35.905 35.907 mp_waitall_1 130650 16.6 27.343 34.822 27.343 34.822 prepare_preconditioner 11 7.9 0.000 0.000 32.063 32.091 make_preconditioner 11 8.9 0.000 0.000 32.063 32.091 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.925 30.179 rebuild_ks_matrix 129 8.3 0.001 0.001 28.853 29.000 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 28.853 28.999 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.425 26.565 multiply_cannon_multrec 10108 15.6 10.456 15.046 18.069 22.586 qs_ot_get_derivative 118 11.6 0.002 0.002 20.292 20.494 multiply_cannon_metrocomm3 10108 15.6 0.022 0.024 13.022 20.471 ot_diis_step 118 11.6 0.020 0.022 20.192 20.193 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.691 20.031 apply_single 129 13.6 0.001 0.001 19.690 20.031 make_m2s 5054 13.6 0.067 0.074 16.618 19.098 make_images 5054 14.6 2.324 2.696 16.315 18.787 cp_fm_cholesky_invert 11 10.9 18.387 18.392 18.387 18.392 qs_ot_get_p 129 10.4 0.001 0.001 18.120 18.289 qs_ot_p2m_diag 83 11.4 0.496 0.501 14.221 14.236 sum_up_and_integrate 129 10.3 0.182 0.192 14.161 14.213 integrate_v_rspace 129 11.3 0.004 0.004 13.978 14.038 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.290 13.343 calculate_rho_elec 129 8.7 0.260 0.272 13.290 13.342 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.075 13.077 make_images_data 5054 15.6 0.054 0.061 10.150 12.795 hybrid_alltoall_any 5240 16.5 0.819 3.838 9.876 12.381 multiply_cannon_sync_h2d 10108 15.6 11.593 12.352 11.593 12.352 init_scf_run 11 5.9 0.000 0.001 11.025 11.026 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.025 11.026 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.030 10.040 cp_fm_diag_elpa_base 83 14.4 9.785 9.850 10.023 10.032 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.161 8.305 multiply_cannon_metrocomm1 10108 15.6 0.030 0.030 4.703 8.168 cp_fm_cholesky_decompose 22 10.9 8.013 8.089 8.013 8.089 grid_integrate_task_list 129 12.3 7.810 8.062 7.810 8.062 pw_transfer 1559 11.6 0.084 0.091 7.902 7.913 dbcsr_mm_accdrv_process 20926 16.1 2.722 3.430 7.245 7.851 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 7.679 7.695 mp_allgather_i34 2527 14.6 2.923 7.650 2.923 7.650 wfi_extrapolate 11 7.9 0.001 0.001 7.272 7.273 calculate_dm_sparse 129 9.5 0.001 0.001 6.664 6.777 fft_wrap_pw1pw2_140 527 13.2 0.506 0.527 6.728 6.748 density_rs2pw 129 9.7 0.005 0.006 6.015 6.258 fft3d_ps 1301 14.7 2.753 2.828 5.818 5.854 grid_collocate_task_list 129 9.7 5.531 5.726 5.531 5.726 dbcsr_complete_redistribute 395 12.7 2.122 2.192 5.282 5.647 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.437 5.446 multiply_cannon_metrocomm4 7581 15.6 0.024 0.027 1.893 5.430 mp_irecv_dv 29086 15.9 1.856 5.348 1.856 5.348 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.220 5.221 mp_alltoall_d11v 2423 14.1 4.824 5.168 4.824 5.168 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.327 4.371 rs_pw_transfer 1054 12.0 0.013 0.014 4.105 4.358 potential_pw2rs 129 12.3 0.026 0.027 4.128 4.142 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.575 3.889 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.790 3.823 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.435 3.752 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.600 3.690 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 3.596 3.641 calculate_first_density_matrix 1 7.0 0.000 0.000 3.557 3.560 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=174.349000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1402.909091, yerr=49.175852 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.042398E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108613552 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3955 59820. MP_Allreduce 10885 1518. MP_Sync 86 MP_Alltoall 1700 36954374. MP_ISendRecv 3556 218624. MP_Wait 11506 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.048 288.442 288.443 qs_mol_dyn_low 1 2.0 0.003 0.004 287.800 287.813 qs_forces 11 3.9 0.005 0.006 287.698 287.703 qs_energies 11 4.9 0.002 0.002 279.121 279.133 scf_env_do_scf 11 5.9 0.002 0.015 256.588 256.598 velocity_verlet 10 3.0 0.002 0.002 207.412 207.421 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 132.259 132.261 init_scf_loop 11 6.9 0.000 0.000 124.053 124.057 prepare_preconditioner 11 7.9 0.000 0.000 119.407 119.427 make_preconditioner 11 8.9 0.000 0.000 119.407 119.427 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.707 116.564 qs_scf_new_mos 116 7.6 0.001 0.001 90.828 90.964 qs_scf_loop_do_ot 116 8.6 0.001 0.001 90.827 90.963 ot_scf_mini 116 9.6 0.004 0.004 86.093 86.152 dbcsr_multiply_generic 2485 12.5 0.214 0.221 83.202 83.668 cp_fm_upper_to_full 104 14.8 52.554 75.182 52.554 75.182 multiply_cannon 2485 13.5 0.700 0.752 58.680 59.985 multiply_cannon_loop 2485 14.5 0.468 0.477 54.407 55.545 ot_mini 116 10.6 0.001 0.001 44.291 44.352 dbcsr_complete_redistribute 393 12.7 3.999 4.064 29.407 41.873 copy_fm_to_dbcsr 208 11.6 0.001 0.001 26.007 38.470 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.653 35.982 mp_alltoall_i22 712 14.1 21.489 34.087 21.489 34.087 cp_fm_cholesky_invert 11 10.9 33.897 33.904 33.897 33.904 rebuild_ks_matrix 127 8.3 0.001 0.001 33.090 33.142 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.018 33.090 33.142 mp_waitall_1 104546 16.7 27.790 31.302 27.790 31.302 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.861 30.899 qs_ot_get_p 127 10.4 0.001 0.001 26.444 26.512 qs_ot_get_derivative 116 11.6 0.001 0.002 24.029 24.095 qs_ot_p2m_diag 82 11.4 0.868 0.872 22.428 22.455 make_m2s 4970 13.5 0.076 0.079 19.761 20.882 cp_dbcsr_syevd 82 12.4 0.006 0.006 20.704 20.706 make_images 4970 14.5 3.766 3.857 19.290 20.415 ot_diis_step 116 11.6 0.022 0.022 20.213 20.214 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.530 19.685 apply_single 127 13.6 0.001 0.001 19.530 19.685 multiply_cannon_metrocomm3 9940 15.5 0.023 0.023 18.354 19.289 multiply_cannon_multrec 9940 15.5 10.396 12.130 17.950 18.045 cp_fm_diag_elpa 82 13.4 0.000 0.000 17.507 17.509 cp_fm_diag_elpa_base 82 14.4 13.212 14.742 17.503 17.505 sum_up_and_integrate 127 10.3 0.317 0.319 15.551 15.640 multiply_cannon_sync_h2d 9940 15.5 15.527 15.535 15.527 15.535 integrate_v_rspace 127 11.3 0.004 0.004 15.233 15.323 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.724 14.749 calculate_rho_elec 127 8.7 0.480 0.481 14.723 14.748 init_scf_run 11 5.9 0.000 0.001 12.570 12.571 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.570 12.570 hybrid_alltoall_any 5155 16.4 1.292 3.021 10.511 12.485 make_images_data 4970 15.5 0.061 0.064 10.504 12.325 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.359 9.414 dbcsr_mm_accdrv_process 20590 16.0 3.691 5.512 7.319 9.135 cp_fm_cholesky_decompose 22 10.9 8.951 9.024 8.951 9.024 grid_integrate_task_list 127 12.3 8.474 8.649 8.474 8.649 wfi_extrapolate 11 7.9 0.001 0.001 8.644 8.645 pw_transfer 1535 11.6 0.090 0.091 8.231 8.240 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 7.999 8.008 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 7.959 7.960 fft_wrap_pw1pw2_140 519 13.2 0.537 0.545 7.048 7.058 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.902 6.962 mp_alltoall_d11v 2401 14.1 6.792 6.929 6.792 6.929 calculate_dm_sparse 127 9.5 0.001 0.001 6.622 6.741 grid_collocate_task_list 127 9.7 6.289 6.326 6.289 6.326 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.072 6.171 fft3d_ps 1281 14.7 2.740 2.759 6.054 6.067 density_rs2pw 127 9.7 0.005 0.005 5.893 5.924 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=288.443000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2697.090909, yerr=191.934390 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/21/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 2.766000E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 419739 0.0% 0.0% 100.0% average stack size 0.0 0.0 22952.9 marketing flops 1.742116E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 1.260077E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 MP_ISendRecv 570 19200. MP_Wait 1302 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.076 86.422 86.423 qs_energies 1 2.0 0.000 0.000 85.948 85.961 ls_scf 1 3.0 0.004 0.029 84.684 84.691 dbcsr_multiply_generic 111 6.7 0.015 0.016 73.342 73.538 multiply_cannon 111 7.7 0.018 0.021 56.513 57.838 multiply_cannon_loop 111 8.7 0.211 0.226 52.977 54.571 ls_scf_main 1 4.0 0.000 0.002 52.436 52.437 density_matrix_trs4 2 5.0 0.002 0.003 46.926 47.009 ls_scf_init_scf 1 4.0 0.000 0.001 29.172 29.174 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.686 27.737 mp_waitall_1 11316 10.9 22.867 26.642 22.867 26.642 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.557 25.574 multiply_cannon_multrec 2664 9.7 8.180 8.918 15.568 17.552 multiply_cannon_sync_h2d 2664 9.7 13.603 15.849 13.603 15.849 make_m2s 222 7.7 0.009 0.012 13.022 13.704 make_images 222 8.7 0.102 0.113 13.000 13.683 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.866 12.352 multiply_cannon_metrocomm3 2664 9.7 0.010 0.012 5.738 8.372 make_images_data 222 9.7 0.005 0.006 7.596 8.297 dbcsr_mm_accdrv_process 4760 10.4 0.519 0.657 7.008 8.225 hybrid_alltoall_any 227 10.6 0.219 1.854 6.572 7.748 dbcsr_mm_accdrv_process_sort 4760 11.4 6.234 7.113 6.234 7.113 calculate_norms 4752 9.8 5.514 6.180 5.514 6.180 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.078 5.286 mp_sum_l 807 5.4 3.296 5.051 3.296 5.051 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.417 3.759 make_images_sizes 222 9.7 0.000 0.000 0.688 3.397 mp_alltoall_i44 222 10.7 0.688 3.397 0.688 3.397 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.071 3.383 mp_irecv_dv 6231 10.9 2.053 3.355 2.053 3.355 arnoldi_extremal 4 6.8 0.000 0.000 3.175 3.204 arnoldi_normal_ev 4 7.8 0.001 0.003 3.175 3.204 ls_scf_post 1 4.0 0.000 0.001 3.072 3.082 build_subspace 16 8.4 0.009 0.012 3.076 3.078 ls_scf_store_result 1 5.0 0.000 0.000 2.899 2.935 dbcsr_special_finalize 555 9.7 0.005 0.007 2.367 2.862 dbcsr_merge_single_wm 555 10.7 0.462 0.598 2.358 2.854 make_images_pack 222 9.7 2.220 2.655 2.222 2.657 dbcsr_sort_data 658 11.4 2.152 2.578 2.152 2.578 dbcsr_matrix_vector_mult 304 9.0 0.004 0.010 2.302 2.544 dbcsr_matrix_vector_mult_local 304 10.0 2.061 2.452 2.063 2.454 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.243 2.331 buffer_matrices_ensure_size 222 8.7 1.765 2.159 1.765 2.159 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.740 1.741 rebuild_ks_matrix 3 7.3 0.000 0.000 1.730 1.732 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.018 1.730 1.732 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=86.423000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1141.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/22/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 5.588524E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 368848 0.0% 0.0% 100.0% average stack size 0.0 0.0 26119.8 marketing flops 1.742116E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 2.089308E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_ISendRecv 282 57600. MP_Wait 828 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.038 90.995 90.997 qs_energies 1 2.0 0.000 0.000 90.546 90.550 ls_scf 1 3.0 0.000 0.000 89.237 89.240 dbcsr_multiply_generic 111 6.7 0.016 0.017 75.432 75.777 multiply_cannon 111 7.7 0.028 0.042 53.420 56.720 ls_scf_main 1 4.0 0.000 0.000 54.837 54.843 multiply_cannon_loop 111 8.7 0.117 0.124 50.125 52.758 density_matrix_trs4 2 5.0 0.002 0.003 49.247 49.441 ls_scf_init_scf 1 4.0 0.000 0.000 30.876 30.878 mp_waitall_1 9246 10.9 21.325 30.109 21.325 30.109 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.722 29.830 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.304 27.319 multiply_cannon_multrec 1332 9.7 13.150 16.644 22.411 27.031 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.964 21.013 make_m2s 222 7.7 0.007 0.008 15.410 16.118 make_images 222 8.7 1.590 1.983 15.380 16.090 dbcsr_mm_accdrv_process 4041 10.4 0.276 0.447 8.864 10.455 dbcsr_mm_accdrv_process_sort 4041 11.4 8.457 10.008 8.457 10.008 make_images_data 222 9.7 0.004 0.005 8.833 9.749 hybrid_alltoall_any 227 10.6 0.523 2.490 8.260 9.059 mp_sum_l 807 5.4 5.294 7.906 5.294 7.906 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.242 7.846 mp_irecv_dv 3311 11.0 3.222 7.793 3.222 7.793 calculate_norms 2376 9.8 6.006 6.805 6.006 6.805 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.165 6.605 multiply_cannon_sync_h2d 1332 9.7 4.746 5.949 4.746 5.949 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.953 5.139 arnoldi_extremal 4 6.8 0.000 0.000 4.589 4.613 arnoldi_normal_ev 4 7.8 0.001 0.005 4.589 4.613 build_subspace 16 8.4 0.014 0.021 4.327 4.330 ls_scf_post 1 4.0 0.000 0.000 3.523 3.525 ls_scf_store_result 1 5.0 0.000 0.000 3.236 3.349 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.132 3.347 dbcsr_matrix_vector_mult_local 304 10.0 2.745 3.215 2.747 3.217 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.161 2.988 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.524 2.611 make_images_pack 222 9.7 2.047 2.463 2.049 2.465 mp_allgather_i34 111 8.7 0.980 2.409 0.980 2.409 dbcsr_sort_data 436 11.2 1.838 2.120 1.838 2.120 dbcsr_data_new 4174 10.1 1.621 1.849 1.621 1.849 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.997000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1736.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/23/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 8.404608E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 353133 0.0% 0.0% 100.0% average stack size 0.0 0.0 27282.1 marketing flops 1.742118E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 2.699223E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_ISendRecv 186 57600. MP_Wait 732 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.049 93.631 93.632 qs_energies 1 2.0 0.000 0.000 92.989 92.992 ls_scf 1 3.0 0.000 0.000 91.598 91.601 dbcsr_multiply_generic 111 6.7 0.016 0.020 76.285 76.610 multiply_cannon 111 7.7 0.040 0.111 53.083 57.883 ls_scf_main 1 4.0 0.000 0.000 57.148 57.165 multiply_cannon_loop 111 8.7 0.101 0.107 49.439 52.751 density_matrix_trs4 2 5.0 0.002 0.003 51.332 51.456 mp_waitall_1 7374 11.0 23.432 32.944 23.432 32.944 ls_scf_init_scf 1 4.0 0.000 0.000 30.845 30.854 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.685 29.759 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.297 27.307 multiply_cannon_multrec 888 9.7 12.657 14.988 21.228 24.215 multiply_cannon_metrocomm3 888 9.7 0.004 0.005 11.051 22.330 make_m2s 222 7.7 0.007 0.008 16.705 17.975 make_images 222 8.7 1.982 2.307 16.666 17.936 hybrid_alltoall_any 227 10.6 0.622 2.859 8.975 10.380 make_images_data 222 9.7 0.004 0.005 9.349 10.348 dbcsr_mm_accdrv_process 3754 10.4 0.259 0.469 8.098 9.384 dbcsr_mm_accdrv_process_sort 3754 11.4 7.715 8.915 7.715 8.915 mp_sum_l 807 5.4 4.920 8.649 4.920 8.649 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.511 7.160 multiply_cannon_sync_h2d 888 9.7 6.063 7.151 6.063 7.151 mp_irecv_dv 2335 11.1 2.496 7.109 2.496 7.109 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.794 6.654 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.655 6.576 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.904 5.163 arnoldi_extremal 4 6.8 0.000 0.000 5.077 5.092 arnoldi_normal_ev 4 7.8 0.001 0.005 5.077 5.091 build_subspace 16 8.4 0.014 0.021 4.767 4.774 calculate_norms 1584 9.8 4.287 4.595 4.287 4.595 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.453 3.769 mp_allgather_i34 111 8.7 1.469 3.731 1.469 3.731 ls_scf_post 1 4.0 0.000 0.000 3.604 3.607 dbcsr_matrix_vector_mult_local 304 10.0 3.033 3.589 3.035 3.591 ls_scf_store_result 1 5.0 0.000 0.000 3.332 3.431 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.723 2.837 make_images_pack 222 9.7 1.843 2.152 1.846 2.155 dbcsr_sort_data 325 11.1 1.903 2.147 1.903 2.147 dbcsr_data_release 9322 10.9 1.326 2.008 1.326 2.008 make_images_sizes 222 9.7 0.000 0.000 0.960 1.907 mp_alltoall_i44 222 10.7 0.960 1.906 0.960 1.906 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.882 1.890 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.632000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2196.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/24/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 10.747127E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 369794 0.0% 0.0% 100.0% average stack size 0.0 0.0 26053.0 marketing flops 1.742116E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 3.385459E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_ISendRecv 138 86400. MP_Wait 600 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.090 0.180 97.980 97.982 qs_energies 1 2.0 0.000 0.000 97.197 97.212 ls_scf 1 3.0 0.023 0.157 95.530 95.539 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.541 78.879 ls_scf_main 1 4.0 0.002 0.027 58.924 59.077 multiply_cannon 111 7.7 0.056 0.125 51.431 55.483 density_matrix_trs4 2 5.0 0.002 0.003 52.876 53.034 multiply_cannon_loop 111 8.7 0.115 0.125 46.312 49.796 ls_scf_init_scf 1 4.0 0.000 0.001 33.258 33.306 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.896 31.974 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.003 29.361 29.371 mp_waitall_1 6438 11.0 22.033 28.351 22.033 28.351 multiply_cannon_multrec 1332 9.7 14.255 17.674 22.177 24.748 make_m2s 222 7.7 0.008 0.009 20.864 22.312 make_images 222 8.7 3.157 3.610 20.813 22.263 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 8.724 17.446 make_images_data 222 9.7 0.004 0.005 11.503 13.127 hybrid_alltoall_any 227 10.6 0.801 3.787 10.854 12.520 dbcsr_mm_accdrv_process 3641 10.4 0.207 0.421 7.570 9.098 dbcsr_mm_accdrv_process_sort 3641 11.4 7.171 8.669 7.171 8.669 mp_sum_l 807 5.4 4.689 8.557 4.689 8.557 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.602 6.715 multiply_cannon_sync_h2d 1332 9.7 5.516 6.313 5.516 6.313 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.115 6.105 mp_irecv_dv 3229 10.9 2.092 6.056 2.092 6.056 arnoldi_extremal 4 6.8 0.020 0.078 5.446 5.458 arnoldi_normal_ev 4 7.8 0.013 0.066 5.426 5.458 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.638 5.020 build_subspace 16 8.4 0.015 0.022 4.975 5.003 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.609 4.810 mp_allgather_i34 111 8.7 2.214 4.692 2.214 4.692 calculate_norms 2376 9.8 4.191 4.540 4.191 4.540 dbcsr_matrix_vector_mult 304 9.0 0.006 0.017 3.648 3.994 dbcsr_matrix_vector_mult_local 304 10.0 3.190 3.684 3.192 3.686 dbcsr_sort_data 658 11.4 3.075 3.511 3.075 3.511 ls_scf_post 1 4.0 0.008 0.047 3.324 3.334 dbcsr_special_finalize 555 9.7 0.006 0.007 2.821 3.276 dbcsr_merge_single_wm 555 10.7 0.536 0.666 2.813 3.268 ls_scf_store_result 1 5.0 0.000 0.000 3.031 3.099 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.876 2.933 dbcsr_data_release 10477 10.7 1.583 2.423 1.583 2.423 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.008 2.062 rebuild_ks_matrix 3 7.3 0.000 0.000 1.984 2.038 qs_ks_build_kohn_sham_matrix 3 8.3 0.114 0.175 1.984 2.038 dbcsr_finalize 304 7.8 0.050 0.062 1.812 2.009 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.982000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2690.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/25/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 15.383312E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 336818 0.0% 0.0% 100.0% average stack size 0.0 0.0 28603.7 marketing flops 1.742118E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 4.628550E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_ISendRecv 90 115200. MP_Wait 573 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.045 0.065 100.361 100.362 qs_energies 1 2.0 0.000 0.000 99.533 99.536 ls_scf 1 3.0 0.000 0.000 97.597 97.611 dbcsr_multiply_generic 111 6.7 0.018 0.018 78.244 78.428 ls_scf_main 1 4.0 0.000 0.000 62.161 62.183 multiply_cannon 111 7.7 0.105 0.233 55.535 61.280 density_matrix_trs4 2 5.0 0.002 0.003 55.153 55.251 multiply_cannon_loop 111 8.7 0.069 0.076 51.109 52.862 mp_waitall_1 5481 11.0 26.839 31.679 26.839 31.679 ls_scf_init_scf 1 4.0 0.000 0.000 31.141 31.155 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.900 29.943 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.579 27.589 multiply_cannon_multrec 444 9.7 14.051 16.344 21.101 22.952 make_m2s 222 7.7 0.006 0.006 17.915 20.303 make_images 222 8.7 3.738 4.437 17.852 20.241 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.133 17.036 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.422 14.187 make_images_data 222 9.7 0.003 0.004 10.085 12.496 hybrid_alltoall_any 227 10.6 0.789 3.756 10.012 12.407 dbcsr_mm_accdrv_process 3003 10.4 0.188 0.346 6.753 7.897 dbcsr_mm_accdrv_process_sort 3003 11.4 6.437 7.549 6.437 7.549 multiply_cannon_sync_h2d 444 9.7 6.545 7.513 6.545 7.513 mp_allgather_i34 111 8.7 2.639 7.033 2.639 7.033 arnoldi_extremal 4 6.8 0.000 0.000 5.786 5.796 arnoldi_normal_ev 4 7.8 0.002 0.005 5.786 5.796 build_subspace 16 8.4 0.015 0.020 5.398 5.408 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.629 4.761 mp_sum_l 807 5.4 2.839 4.624 2.839 4.624 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.702 4.475 mp_irecv_dv 1241 11.2 1.680 4.434 1.680 4.434 dbcsr_matrix_vector_mult 304 9.0 0.007 0.017 4.143 4.340 ls_scf_post 1 4.0 0.000 0.000 4.295 4.298 dbcsr_matrix_vector_mult_local 304 10.0 3.678 4.138 3.680 4.140 calculate_norms 792 9.8 3.561 3.694 3.561 3.694 make_images_sizes 222 9.7 0.000 0.000 1.153 3.533 mp_alltoall_i44 222 10.7 1.153 3.532 1.153 3.532 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.374 3.462 ls_scf_store_result 1 5.0 0.000 0.000 3.375 3.409 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.901 3.384 dbcsr_finalize 304 7.8 0.062 0.078 2.209 2.277 dbcsr_merge_all 275 8.9 0.474 0.525 2.054 2.100 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.021 2.024 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=100.362000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3642.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8c0411e32bbb6770871f8515127d8b2d742a8838_performance_tests/26/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 30.358840E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 339931 0.0% 0.0% 100.0% average stack size 0.0 0.0 28341.7 marketing flops 1.742118E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 8.726553E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_ISendRecv 84 732600. MP_Wait 309 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.079 0.099 107.211 107.211 qs_energies 1 2.0 0.000 0.000 105.826 105.840 ls_scf 1 3.0 0.000 0.000 102.942 102.955 dbcsr_multiply_generic 111 6.7 0.023 0.027 77.038 77.197 ls_scf_main 1 4.0 0.000 0.000 65.358 65.372 density_matrix_trs4 2 5.0 0.002 0.003 56.544 56.605 multiply_cannon 111 7.7 0.150 0.226 48.950 50.096 multiply_cannon_loop 111 8.7 0.068 0.071 45.508 47.008 ls_scf_init_scf 1 4.0 0.000 0.000 33.949 33.963 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.616 32.635 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.784 29.789 mp_waitall_1 4569 11.1 21.233 25.476 21.233 25.476 make_m2s 222 7.7 0.007 0.007 23.938 24.867 make_images 222 8.7 4.613 5.036 23.829 24.755 multiply_cannon_multrec 444 9.7 17.870 18.657 22.493 23.071 hybrid_alltoall_any 227 10.6 1.659 3.616 12.945 15.576 make_images_data 222 9.7 0.004 0.004 13.118 15.490 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 9.538 10.367 multiply_cannon_sync_h2d 444 9.7 8.851 8.894 8.851 8.894 arnoldi_extremal 4 6.8 0.000 0.000 7.466 7.474 arnoldi_normal_ev 4 7.8 0.003 0.009 7.466 7.474 build_subspace 16 8.4 0.026 0.036 6.920 6.933 dbcsr_matrix_vector_mult 304 9.0 0.009 0.026 5.467 5.626 dbcsr_matrix_vector_mult_local 304 10.0 5.034 5.352 5.036 5.354 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.036 5.290 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.749 4.841 dbcsr_mm_accdrv_process 1814 10.4 0.212 0.316 4.454 4.596 dbcsr_mm_accdrv_process_sort 1814 11.4 4.147 4.278 4.147 4.278 ls_scf_post 1 4.0 0.000 0.000 3.635 3.649 make_images_sizes 222 9.7 0.000 0.000 1.440 3.595 mp_alltoall_i44 222 10.7 1.440 3.595 1.440 3.595 ls_scf_store_result 1 5.0 0.000 0.000 3.382 3.398 calculate_norms 792 9.8 3.240 3.272 3.240 3.272 mp_allgather_i34 111 8.7 1.052 3.204 1.052 3.204 dbcsr_finalize 304 7.8 0.082 0.090 3.087 3.155 mp_sum_l 807 5.4 1.817 3.073 1.817 3.073 dbcsr_merge_all 275 8.9 0.895 0.923 2.872 2.934 dbcsr_complete_redistribute 5 7.6 1.441 1.480 2.759 2.876 qs_energies_init_hamiltonians 1 3.0 0.002 0.002 2.854 2.854 dbcsr_data_release 12724 10.6 2.330 2.840 2.330 2.840 matrix_ls_to_qs 2 6.0 0.000 0.000 2.405 2.531 dbcsr_sort_data 325 11.1 2.437 2.514 2.437 2.514 dbcsr_new_transposed 4 7.5 0.243 0.253 2.294 2.314 dbcsr_frobenius_norm 74 6.6 2.057 2.133 2.211 2.256 dbcsr_add_d 103 6.2 0.000 0.000 2.137 2.213 dbcsr_add_anytype 103 7.2 0.860 0.893 2.136 2.213 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.189 2.204 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.211000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6822.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 8c0411e32bbb6770871f8515127d8b2d742a8838 Summary: empty Status: OK